7/29/2019 1922 - Handwriting

1/158

\ Wv- :'' '.v; r- : ' I' 1 .', / i . ' .' : - ''-: .

;

BS88S5HBi^5r^w3Ss' wist' ' BeS* BJssi aHSsp ,#w liwSfc Mffsf&ramfer" ' ' '

:

'

7/29/2019 1922 - Handwriting

2/158

THIS BOOK IS THE GIFT OFHenry Pelouze de Forest

Class of 1884

7/29/2019 1922 - Handwriting

3/158

Ctittf) Jftacomfcer f|ali(iw. jfranklin J^all)

Phone2430. 79TH ST.New York City

JBtpartment e,'W ftUlfaattt"154-160 east 9 ut stmt

&eto Horfe dtp,

(draptjolog^ Can Benefit 3gouGraphology reveals personal weakness and moral defects,

genius or mediocrity, enabling you to cultivate your talents, andavoid pitfalls, by suppressing your faults. It is a mirror reflectingthe trustworthiness of friends or business associates; the congenial-ity of those seeking life mates.A graphologist cannot tell your fortune; but as truly aschemistry reveals the component parts of liquids and minerals,showing the destructiveness of certain combinations, the benefit ofothers, so graphology scientifically characterizes you from yourhandwriting, teaching that if your faults, pettiness, sensuality,selfishness and pessimism predominate, you will probably be ship-wrecked upon the sea of Life ; while, if you possess more good thanevil traits, intelligence, force, energy, ambition and honor, withpower to conserve and concentrate your forces, you will become" Master of your Fate."

One whose expert work as a reader of character has won un-stinted praise from the press and people of many lands is Mrs.Franklin Hall, M.S.G. The degree, " Master of ScientificGraphology " having been conferred upon her by a well knownwestern college, several years ago.

Mrs. Hall is a descendant of scientists and scholars whosenames are revered by students and philosophers. On her father'sside she traces her lineage directly back to one of the nobles whowas a trusted ally of Robert Bruce ; the crest upon the family ,coat-6f-arms having been bestowed upon this ancestor for havingsaved the life of the Bruce.A pupil of the once famous San Souci Seminary, Ballston Spa,N. Y., Mrs. Hall early developed a love for science and literature.For several years she held important editorial positions, until hersuccess as a delineator of character from handwriting made suchexacting demands upon her time that she was obliged to limit herliterary work to occasional contributions of poems, stories andarticles to well known papers and magazines. Her fame as a

7/29/2019 1922 - Handwriting

4/158

graphologist has penetrated even into " Darkest Africa ; " in fact,into different parts of the globe where little English is spoken, thusnecessitating translation.

During the Roosevelt-Parker campaign, Mrs, Hall was hon-ored by a request from a prominent syndicate, for delineations ofcharacter from the writing of the rival candidates. These wereso favorably; received that they were copied and given prominenceby all of the leading newspapers of the country.

The following 'are a few of the American newspapers whichhave (given generous space to personal articles and interviews re-garding Mrs. Hall's achievements in science and literature.

Leslie's Weekly, New York Herald, Chicago Times-Herald(now Record-Herald), Success, Opportunity, New York EveningTelegram, Los Angeles Times, The Troy Press, Hartford Globe,Troy Daily Times, Pittsburg Post, Toledo Blade, RochesterDemocrat, Indianapolis Star, etc., etc. To the above list couldbe added thousands of personal testimonials, were there space.A gentleman writes from Wiesbaden, Germany: "Yourname is as well known on this side of the Atlantic as at home.Not to have had a reading of character from Mrs. Hall, is to beout of date."E. L. "

One of the most rioted philologists from Paris, France:" Mrs. Hall, your work is marvelous; your readings are of price-less value to the seeker after knowledge."Prof. J. Sulzbache.From the Alps in Switzerland: " I am resting here; dosend me one of your character delineations from the enclosed slip,the writing of a wayfayer in whom I am interested. I trust noone until you have analyzed them for me."Mrs. S. R.

From London : " I have always laughed at the folly of any-one being able to decipher character from handwriting. I am nolonger a skeptic; Mrs. Franklin Hall has convinced me it can beaccurately done."F.

"Mrs. Hall, permit me to thank you; your readings havebenefited me mentally and physically and made me a better andstronger man."F. T. G, Hawaii.

" You startled me and compelled me to turn over a new leaf.Yes, I am untruthful and possessor of the other bad traits men-tioned. If I become worthy in the future, to youone whom Ihave never seenbelongs the praise."T., Milan, Italy.If you, like the writers quoted, desire to learn more of self,write thirty words or more with pen and ink upon unruled paper,enclosing necessary fee. Disguised writing is valueless. Sex isnot always discernible in writing, owing t,o the present day mascu-line' mentality of many women arid effeminacy of an equal pro-portion of men.

7/29/2019 1922 - Handwriting

5/158

jfees for fteatitnpCASH MUST ACCOMPANY all orders.STAMPS NOT ACCEPTED in payment.Price of readings not dependent upon length but time spent

in careful analysis.Readings 25c, 50c, $1.00 and $2.00. Personal readings, byappointment only, $5.00.Ten cents must be added to all checks from outside NewYork City, to cover cost of collection.Personal attention given to all readings, and strictest con-

fidence assured.Special rates for party, club or dinner orders.

3tt Business or Social 3Life igou j&eefc" The Brain Behind the Pen," a book just issued by Mrs. Frank-lin Hall, consisting of twelve lessons fully illustrated, made clearand concise, to enable you to read character from the handwritingto judge an employee or a friend or acquaintance. A charminggift book, beneficial and fascinating in its interest. Price $1.50.Outside the international postal limit, $1.60.

Self-addressed envelope enclosed, which please use in all cor-respondence to insure against misdirection.

fteatrings for JfatrsA decided money maker. Mrs. Hall gives these readings

either upon a cash or commission basis. Write for particulars.e

! A lady writing from Lakewood, where Mrs. Hall gave read-ings for four days, says : " We can never sufficiently thank youfor your work at the Hospital Fair given by our society ladies.Your booth, a magnet for all, was a winner financially."

From three days at a church fair in Brooklyn : " We clearedmore from your little room set aside for Graphology, which wascontinually crowded, than from anything else. Thanks."

After an Orphan's Fair, New York City: "We are un-ceasingly grateful for the large sum realized from your characterreadings at the Tuxedo."

Make all checks, postal and express orders payable to

jftrs* Jfranfclin flail160 E. 91st St. New York City

7/29/2019 1922 - Handwriting

6/158

Si

7/29/2019 1922 - Handwriting

7/158

7/29/2019 1922 - Handwriting

8/158

The Classificationand Identification ofHANDWRITING

By C. D. LEE and R. A. ABBEY

Handwriting, though it is a principalfactor in many criminal cases, and thoughit has been the subject of expert study formany years, has not yet reached the im-portance which finger prints have as-sumed in criminology.

This is solely for the reason that itstreatment has not reached the samescientific level. The classification ofhandwriting, the handling of specimensand files, has not kept pace with the studyof characteristics and with skill in iden-tification.

In the Police Department of Berkeley,California, C. D. Lee, Superintendent ofRecords, and R. A. Abbey, HandwritingExpert, have worked out a scientific andthoroughly practical system for classify-ing and filing specimens of handwriting.This they are using with excellent effectin their actual work.This system is extremely simple. The

proper principle once determined, theworking out of the details follows natu-rally and without complication. Thusthe system possesses the essential featurethat it is not difficult to understand, tonstall, or to operate.In this volume Messrs. Lee and Abbey

iescribe their system carefully. They

=intended it either a pled)

NAME

rtgage of the above described p

Name...Residence.....j?...

NAMEMAddress

ExtendedName..JtesideWe c^LClW 68BIB.

NAME,.... U-~*Residence J...JL.IJ

UPPER: Two sigperson. According ireceived the same clwere readily establisl

LOWER: Four s:same person. When ication system descrilvaried only in the21,221,132 or 3. It *tures were all by the

7/29/2019 1922 - Handwriting

9/158

tgage of the above described

as an absolute sale.

...Mm..

!MMExtended..

yvLttxu

pawnbooks, by the same:m outlined here, thesenumber, 31,221,131, andrork of the same hand.

i pawnbooks, all by thetudied under the classifi-le classification numberle eight figures, beingtermined that the signa-

show exactly how specimens of hand-writing are classified, and how they are

;!

filed. They show how specimens of un-'known authorship, after being classified, ;may speedily be checked with knownhands for identification, -and how theirclassification and filing methods renderthis process almost effortless.

Since classification is largely for the \purpose of aiding identification, thesection of this book dealing with indenti-fication is almost as important and fullyas interesting as that describing classifi-cation and filing. The authors tell just

.

how handwriting is studied by experts,just what characteristic details may be idiscovered, and just what may be deter-mined from scientific scrutiny of a speci-men of handwriting. This portion of thebook embodies the methods and ideas ofleading handwriting authorities such asOsborn, Ames, Hagen, Frazer and Lo-card. This discussion is of great im-

jportance in criminology, and will proveimmensely valuable and interesting toeveryone concerned at all with analysis ',of handwriting for any purpose.The book is abundantly illustrated

with reproductions of specimens showingvarious types of handwriting and the pro-cesses by which handwriting is identifiedand classified. These are valuable andinteresting in themselves, and serve ad-mirably to supplement and clarify thetext.

8vo. Cloth BindingPrice #2.50 net

IllustratedBy Mail #2.65

7/29/2019 1922 - Handwriting

10/158

The Classification and Identificationof Handwriting

classification Iblr Eyes*> 21.211,111 31k mar

7/29/2019 1922 - Handwriting

11/158

Cornell UniversityLibrary

The original of this book is inthe Cornell University Library.

There are no known copyright restrictions inthe United States on the use of the text.

http://www.archive.org/details/cu31924032597480

7/29/2019 1922 - Handwriting

12/158

7/29/2019 1922 - Handwriting

13/158

CLASSIFICATION ANDIDENTIFICATIONOF HANDWRITING

7/29/2019 1922 - Handwriting

14/158

7/29/2019 1922 - Handwriting

15/158

CLASSIFICATION ANDIDENTIFICATIONOF HANDWRITING

BYC. D. LEE

SUPERINTENDENT OF RECORDS, POLICE DEPT., BERKELEY, CALIFORNIAAND

R. A. ABBEYHANDWRITING EXPERT, POLICE DEPT., BERKELEY, CALIFORNIA

D. APPLETON AND COMPANYNEW YORK ":: 1922^ :: LONDON

7/29/2019 1922 - Handwriting

16/158

J;r V

COPYRIGHT, 1922, BYD. APPLETON AND COMPANY

PRINTED IS THE UNITED STATES OF AMEBICA

7/29/2019 1922 - Handwriting

17/158

v'Tjjpsa*-'

PRINTED IN THE TOUTED STATES 01- AMERICA

j liR 3

/^

7/29/2019 1922 - Handwriting

18/158

A Character TestinA GameERE'S an amusement feat-ure for yourself and guests."Soul-searching" pastimesare ever the favorites,aren't they? Having for-

tunes told, horoscopes read, tealeaves studied for the handsomedark gentleman and the muchpromised legacy always insure a gayhour.This is a horoscope game, bywhich you may find out underwhich sign of the zodiac your "vic-

tim" was born, and how much heknows about himself as the "stars"know about him.Because I have given five ques-

tions for each of the twelve zodiacalsigns with five outstanding traitssupposed to belong to it, you havesixty queries covering the entirezodiacal year. The zodiacal signs,you understand, do not covermonths from the first day thereofto the 30th or 31st. They have theirown system of time recording, asyou will note beneath each groupof questions.

In order to shorten the tests, ifyou have a number of guests to bequestioned, divide them into groupscovering six month periods. Thatis, take first those who admit beingborn between the 21st of March andSeptember 23. Then take the othergroup born between September 23and March 21 for questioning. Oryou may divide them further intoquarterly divisions of the zodiacalyear, say those with birthdays com-ing between March 21 and June 21.Between June 21 and September 23.Or, if' you are trying to determinethe birth month of one individualthis three month division is best.All the questioned one is to admitis that his birthday comes duringthat three month period. He is notto reveal the exact date until thequestioning is completed.

If the one queried admits "yes" inthree answers out of the five in hisown sign you may believe he has apretty good line on his own charac-only incidental, really, to the pleas-anter occupation (to most of us) ofindulging in character analysis. Allthe queries in each group should beanswered by "yes" to agree with thehoroscope character readings. But,of course, none of it is to be takentoo seriously.

QuestionsDoes your head govern your ac-

tions entirely?Are you strongly independent?You must do everything in yourown way?You can be controlled throughyour reason?You love music and dancing?Aries (21st of March to April 19)

is the sign under which you wereborn, for these are the outstandingtraits of the sign.

Are you "stiff-necked," unyield-ing?

Signs of the ZodiacGive Lines onYour Nature

Are you' in danger of extremes ineverything you do? One minutebold and aggressive, the next timidand conservative?That's the Gemini bornfromMay 20th to June 21st.Are you very domestic? Lovehome and family?Are you overanxious about mak-

ing money and laying up for thefuture?Are you industrious and econom-ical to the extreme?Are you exceedingly sensitive tophysical and mental conditions ofothers?Are you maternal in your rela-tions to people? Even though theyare men?Cancer is your sign. From June21 to July 22 are the dates it in-cludes.

Can you readily adapt yourself toany condition of life?

Virgo is your sign. Prom August22d to September 22d.Are you positive and decisive in

everything you do?Are you inclined toward spiritual-ism or anything psychic?Are you demonstrative in yourlove?Are you inclined to periods ofmelancholy?Annoyed if asked for a reason foryour decisions and actions?That's Librafrom September 23

to October 23.Have you a high temper?Are you jealous?

,Are you disinclined to work withyour hands?Are you inclined to be extrava-gant and self-gratifying?Do you want to be boss?Scorpio is your sign. October 23

to November 22 are the dates itcovers.Are you bold, fearless, determined,and combative?Are you quick to decide, act, andspeak?Are you a person of one thoughtand one idea at a time?Is your mind constantly beyond

the present, even to such small1 things as announcing who is corn-

Are you more controlled by intui-tions than reasoning?Are you easily led or misled byfriends?Do you jump at "and act from con-clusions too quickly?

ing when the doorbell rings or foot-steps are heard?Do you hate to give in whenwrong?Sagittarius is yonr sign. Nov 22-to Dec. 21.

7/29/2019 1922 - Handwriting

19/158

DEDICATED TOTHE INTERNATIONAL ASSOCIATION

FOR IDENTIFICATION

7/29/2019 1922 - Handwriting

20/158

7/29/2019 1922 - Handwriting

21/158

PREFACETwenty years ago a system for the classification of

fingerprints was devised by Sir E. R. Henry, Commis-sioner of Police of London. Many years before the ad-vent of the Henry system scientists had discovered thevalue of fingerprints as a simple and certain means ofidentification, but it remained for a man acquainted withpolice problems to perfect a scheme for the systematicfiling of thousands of these records so as to be readilyavailable for identification purposes at all times.

The position of the A-L System of Classificationand Identification of Handwriting is to-day analogousto that of the Henry system twenty years ago. Thatin every one's handwriting certain definite and fixedcharacteristics exist which serve to establish the writer'sidentity has long been recognized, but to date nothinghas been published in this country on the subject oftheir classification. Dr. Hans Schneickert of the Metro-politan Police of Berlin describes a system for theclassification of handwriting in the "Archive for CourtInvestigations of Handwriting." His system is adaptedto the German script and is divided into two parts, theprimary classification being based upon a crime cate-gory and the secondary upon graphic characteristics.

vii

7/29/2019 1922 - Handwriting

22/158

PREFACEIn the present system a crime category has been

dispensed with in order that a single search of the filesmay suffice to establish whether or not a duplicate existstherein, which is believed to conform more closely to themodern idea of centralization of records. Thus, as withfingerprints, the writing specimen itself contains all theelements necessary for its correqt classification, regard-less of the criminal proclivities, name, description, etc.,of the author. A thief to-day may turn forger to-mor-row or blackmailer the next day; so were his specimenfiled first according to criminal specialty, and secondaccording to the graphic characteristics of his writing,search would be necessary under a number of crime di-visions in order to make his identification certain.A great amount of research work covering thousandsof signatures was found necessary in perfecting the pres-

ent system. The authors are indebted to their chief,August Vollmer, at whose solicitation the work wasundertaken, for his many helpful suggestions; also toOsborn's Questioned Documents, Ames on Forgery,Frazer's Bibliotics, Hagan's Disputed Handwriting,Blackburn, Douglas, Caddell, and Warthman on TheDetection of Forgery, and others, without which theirefforts would have proved futile. Many factors testedfor classification purposes were found wanting in someparticular and had to be abandoned; while with othersthat were finally adopted it was found necessary tochange their relative positions on the chart with a view

viii

7/29/2019 1922 - Handwriting

23/158

PREFACEto giving first place in the classification to those factorsleast susceptible of change or disguise.No system will ever supplant fingerprints for the

purpose of direct personal identificationthat is, forthe identification of the man under arrest. But for in-direct identification, that is, for the fixing of guilt uponthe unknown offender, in the commission of whose crimehandwriting constitutes the corpus delicti, it is the beliefof the authors that an efficient system for the classifi-cation and identification of handwriting will surpass inusefulness the fingerprint system.

However, as between the two systems, there is thisessential difference, that, whereas in the latter theminutiae, or the elements upon which identification de-pends, are unchanged from birth to death exceptthrough accident, in handwriting many of the identify-ing characteristics are susceptible of change by design,and some show natural changes due to fatigue, sickness,age, etc.

So infallibility for any system of handwriting classi-fication can never be attained; but society will be ren-dered a real service if the present system succeeds inbringing to justice an occasional transgressor of itslaws.

IX

7/29/2019 1922 - Handwriting

24/158

7/29/2019 1922 - Handwriting

25/158

ACKNOWLEDGMENTThe authors and publishers desire to give special thanks to Mr.Albert S. Osborn for his work, Questioned Documents, whichhas been of invaluable service in preparing this volume, Classifi-cation and Identification of Handwriting. The following sectionscontain material quoted from or based upon Mr. Osborn's work.Page

7/29/2019 1922 - Handwriting

26/158

7/29/2019 1922 - Handwriting

27/158

CONTENTSCHAPTEB PAGE

Preface vii

INTRODUCTION = XVI. Collecting Material fob the File .... 1

II. Fundamentals of the System 5III. Classification 9

IV. Indexing and Filing 34V. Identification 40

Bibliography - HIIndex 113

7/29/2019 1922 - Handwriting

28/158

7/29/2019 1922 - Handwriting

29/158

ILLUSTRATIONSPAGE

Fig. 1A.Illustrations of Factors I, II, III, IV . .10Fig. IB.Illustrations of Factors V, VI, VII, VIII. . 11Fig. 2.Illustrating the Three Classes of (I) Form . 15Fig. 3.Illustrating the Three Classes of (II) Skill . . 17Fig. 4.Illustrating the Three Classes of (III) Connections . 19Fig. 5.Illustrating the Three Classes of (IV) Shading . 21Fig. 6.Illustrating the Three Classes of (V) Movement . 25Fig. 7.Illustrating the Three Classes of (VI) Embellish-

ment . . 27Fig. 8.Illustrating the Three Classes of (VII) Terminals 29Fig. 9.Illustrating the Three Classes of (VIII) Slant . . 31Fig. 10A.Illustrating Faulty Alignment . . 48Fig. 10B.Illustrating Faulty Alignment . ... 49Fig. 11A.Illustrating Uniformity of Angle Values . 50Fig. llB.Angle Value Test used to Prove Forgery . . 50Fig. 11C.Showing Different Signatures . . 51Fig. 11D.Comparing Checks to Detect Forgery . . 52Fig. HE.Showing Various Angles . . . 53Fig. 12A.Characteristic Dash after Name and Amount . 56Fig. 12B.Proximity of Writing to Printed Headings . 57Fig. 13.Averages and Divergencies ... . . 60Fig. 14ACharacteristics 62, 63Fig. 14B.Trade-marks on Checks for Expressing Cents . 64Fig. 15.Curves . . . 65Fig. 16.Line Quality . 76pIG 17,Muscular Habits and Pen Position . . . 82

xiii

7/29/2019 1922 - Handwriting

30/158

ILLUSTRATIONSPAGE

Pig. 18.Retrace 91Fig. 19.Spacing 93Fig. 20.System . . 94Fig. 21.Unquestioned Identity . .... 97Fig. 22.Checks Trade-marked with Double Dots under

Fraction Lines and Signature . ... 98Fig. 23.Checks not Conspicuously Trade-marked 99Fig. 24.Two Pawnbook Signatures Above and Exemplars

Below . . 100Fig. 25.One Pawnbook Signature Above with Exemplar

Below .100Fig. 26.Forged Indorsements with Exemplar of Forger

Below . 101Fig. 27.Two Worthless Checks and Standard Writing of

Author 102Fig. 28.Two Checks Written by Same Person . 103Fig. 29.Two Worthless Check and Standard Writing of

Author .... 104Fig. 30.Attempt at Forgery by Feeble-minded Boy of

Twenty-one . . . 105Fig. 31.Two Signatures on Pawnbooks by Same Person . 105Fig. 32.Four Signatures on Pawnbooks by Same Person . 106Fig. 33.Blackhand Letter, Author Unknown . . . 107Fig. 34.Conspicuous for its Large Eyelets .... 107Fig. 35.Classification ... ... 108Fig. 36.Specimen Cards and Index Cards Reduced . . 110

XIV

7/29/2019 1922 - Handwriting

31/158

INTRODUCTIONA great amount of handwriting material in the form

of checks, fraudulent, fictitious and forged; blackmail-ing letters, threatening and anonymous letters, lettersfrom beggars and swindlers, is accumulated each yearby the police, who have been handicapped through lackof adequate means for filing it, except numerically bythe case number, or alphabetically under the name of thewriter, if known, or of the recipient, or perhaps by theparticular crime. Under these conditions, when a docu-ment of unknown authorship came under policescrutiny, the only hope of identifying the writer lay inobtaining the services of a handwriting expert, whomust go through the thousands of specimens in the filesin order to attain his end. The inordinate amount oftime and labor required to do this often precludes thepossibility of any practical results being obtained in theway of identification, as a result of which there has longbeen felt among identification experts and policeauthorities generally the need of a system for classify-ing and filing this material.

The purpose of the A-L System is to provide anorderly means for the filing of specimens of handwrit-ing, whether a single signature or a voluminous docu-

xv

7/29/2019 1922 - Handwriting

32/158

INTRODUCTIONment, so that should the same writing come into questionat any future time, it may be found and the author iden-tified in much the same manner as is done with finger-prints. By means of this system and with the knowl-edge of handwriting identification to be gained by astudy of this work, any one can readily locate in a filethe duplicate of any specimen at hand provided it existstherein; or, with equal ease, place the specimen in thefile under its proper classification for future reference.

At the time of writing, the system has been in actualoperation in the Police Department at Berkeley, Cali-fornia, for three months, where practical results havealready been obtained in the way of several importantidentifications; and the system has just been installedby the California State Bureau of Identification andInvestigation at Sacramento.

xvi

7/29/2019 1922 - Handwriting

33/158

CLASSIFICATION ANDIDENTIFICATIONOF HANDWRITING

7/29/2019 1922 - Handwriting

34/158

7/29/2019 1922 - Handwriting

35/158

CLASSIFICATION AND IDENTIFICATIONOF HANDWRITING

CHAPTER ICOLLECTING MATERIAL FOR THEHANDWRITING FILE

In the ordinary police procedure when a person isarrested for a serious offense his photograph and finger-prints are taken for the purpose of present and futureidentification. For the purpose of the handwriting col-lection the prisoner is now required to sign, with amedium flexible pen, in his natural hand, a specimencard at the time he is fingerprinted. An 8x5 inchcard is used, which may be white, or a color scheme maybe utilized to designate sex and race as explained inChapter IV. Headings may be printed on the card asfollows

Classification

.

Hair Eyes Height Weight Age

Name in fullAddressCityOccupation Date of birth(To be written by prisoner)Crime Case Number

.

Record Number Date signed . .Modus Operandi(To be filled in by police)

7/29/2019 1922 - Handwriting

36/158

CLASSIFICATION OF HANDWRITINGIn due course this specimen card is given to the

handwriting classifier, who classifies, searches and filesit in its proper place in the collection. In the course ofthe search, former writings of the same person may beturned up in the form of a forgotten check or athreatening letter; whereupon the system is creditedwith an identification and an additional charge may befiled against the defendant.

Another source from which our material may be col-lected is the writing of various kinds, as already sug-gested, which is submitted to or taken by the police asevidence. This may be pasted or otherwise fastened tothe specimen card or, if bulky, placed in a 8 x 5 folderfor filing, as explained in Chapter IV. While the ma-terial derived from this source is usually of unknownorigin, it is treated in exactly the same manner as thatjust described where the author is known. And here,too, the search may result in a speedy clearing up of thecase. When such evidence is to be used in court andthere may be difficulty in securing it later for the col-lection, it should be photographed before leaving thehands of the police ; but, of course, the best material forour purpose is the original writing itself.

Other material is derived from the numerous policecirculars containing copies of checks and other fraudu-lent documents submitted by other departments. Theseare treated as before, and when an identification is madethe issuing department is notified thereof.

2

7/29/2019 1922 - Handwriting

37/158

COLLECTING MATERIALThe police bureaus of identification receive many

duplicate fingerprint records of the same man, whichordinarily bear his signature, and the signature may beclipped from these, pasted on a specimen card and filedin the regular way.

From whatever source a specimen of handwriting isderived, it represents the work of either a known or anunknown writer. The purpose of filing writings ofknown origin is to make possible the identification offuture writings by the same author under whatever cir-cumstances of modus operandi or name. Therefore, weshould make certain, as far as possible, that the specimento be filed is in the natural hand of the writer ; and newspecimens should be obtained for the files at each sub-sequent arrest if a period of more than four years haselapsed.

The object of filing writings of unknown origin isobvious, that is, the discovery of the author and theclearing up of the case. If the search fails to reveala duplicate of the questioned writing in the files,then the specimen is filed in the hope of future identi-fication.

Several specimens of the same unknown writer maybe received from different sources at different timesand brought together in the files. In such case thespecimens may be clipped together or they may be filedseparately- under different classifications as reference;

S

7/29/2019 1922 - Handwriting

38/158

CLASSIFICATION OF HANDWRITINGcards when the necessity arises, as explained in ChapterIII, where they will remain in the files until the writeris identified. It should be understood at the outset thatthe more writing material available in a given case, themore correct classification and positive identification arefacilitated.

7/29/2019 1922 - Handwriting

39/158

CHAPTER IIFUNDAMENTALS OF THE SYSTEM

A great deal of the writing material collected by thepolice consists of a mere signature, such as found inpawnbroker's books, hotel registers, fraudulent endorse-ments on stolen checks, etc. ; so any system of classifica-tion to be successful for general police use should beworkable with even so small an amount of material asa single signature. A single signature may consist ofas few as two capitals and two small letters. Thereforethe A-L System is based on general writing character-istics found in any combination of letters rather thanon individual characteristics occurring in certain lettersonly.

In the process of perfecting the present system thedifficulty of selecting suitable criteria for classifying asingle signature became at once apparent. Thus intesting the most significant of the handwriting char-acteristics, it was found that many of them could not beapplied in any manner to the classification of a meresignature. For instance, a single signature in themajority of cases would fail to furnish any index topossible faulty alignment; and the same is true ofarrangement; any system of measurements and aver-

3

7/29/2019 1922 - Handwriting

40/158

CLASSIFICATION OF HANDWRITINGages to be of value must be based upon a considerableamount of material; evidence of muscular incoordina-tion does not manifest itself in classifiable form in allwritings, which is to be regretted, as this is one of themost difficult of writing characteristics to be concealed,because one of the most unconscious and inconspicuous,and often quite beyond the power of the writer to con-trol.

So that where a considerable amount of writingmaterial is available in a given case many factors areapplicable for purposes of classification, in a singlesignature we find ourselves limited to but a few. It isthe belief of the authors that in the present system thisdifficulty is completely surmounted.

Referring to the Classification Chart, Fig. 1, it isseen that eight factors are utilized, each of which isdivided into three classes. The fundamental principleunderlying this tripartite scheme is that each factor isconsidered from the viewpoint of two extremes, or, itmight be termed in some cases, two departures fromthe normal, i.e., classes 1 and 3; and that where thespecimen under consideration does not fall clearly intoone or the other extreme, it is placed in the middle orClass 2. Generally speaking, the occurrence of a singlecharacteristic in a handwriting specimen does not war-rant its classification in accordance therewith, but ratherthe general appearance of the writing as a whole is theguide,

6

7/29/2019 1922 - Handwriting

41/158

FUNDAMENTALS OF THE SYSTEMThus, in the matter of skill, the appearance of a

singleprobably accidentalmisshapen letter wouldnot justify the classification as poor (Class 1) ormedium (Class 2) of an otherwise legible, symmetricaland pictorially pleasing specimen; nor a single super-fluous stroke warrant its classification as an embellishedwriting.

With the specimen card before us we study it firstas to its Form. If we determine from the explanationsand rules set forth in Chapter III that it is an angularwriting, we place a figure 1 at the top of the card afterthe heading "Classification." Next we take up thesecond factor and examine it as to the degree of Skillwith which it was written, and we may find all the ele-ments of a good or medium hand lacking ; so we classifyit as poor and place a figure 1 after the 1 already noted.Next, examined as to Connections, we find several ofthe small letters disconnected, so we add a figure 3 toour classification number. And so we proceed throughthe eight factors, determining the class of each, untilwe have a classification number consisting of eight digits,under which number the specimen is filed away as ex-plained in Chapter IV.

It is to be noted that the determination of the classof five of the factors, Form, Skill, Movement, Embel-lishment and Terminals, does not require the use of anyinstrument whatever. In the matter of Connections,lines may pccasjopally have to be examined under some

7

7/29/2019 1922 - Handwriting

42/158

CLASSIFICATION OF HANDWRITINGmagnification to detect overlapping resulting from lift-ing the pen and starting a new line over the end of thedisconnected stroke. Shading may sometimes requirescrutiny and measurement under low magnification, andthe degree of slant is determined by the use of a trans-parent protractor. Thus the system of classificationpresents no difficulty in the way of technical or com-plex procedure; and we are confident that once certainstandards are firmly fixed in the mind of the classifier,it will prove very simple and highly efficient. It isimportant, however, that the classifying all be done asfar as possible by one person who will specialize in thiswork, since here, as with fingerprints, opinions mayoccasionally differ as to the classification of certaintransitional types, and efficiency demands consistency insuch cases.

By way of practice in classifying, it is suggestedthat the student secure a number of different signatureson small cards and number each card. Then on aseparate sheet of paper opposite the card number writethe handwriting classification of each. Repeat thisoperation at intervals and compare the results with thosefirst obtained, and in a short time it will be found thatthe classification number will always be the same. Thisprocedure will develop confidence and will be of assist-ance in establishing in the mind of the classifier thestandards necessary to success.

8

7/29/2019 1922 - Handwriting

43/158

CHAPTER IIICLASSIFICATION

I. FormWhat is here meant by Form is the general appear-

ance of the writing from the viewpoint of angularity orrotundity as seen in the arcs of such letters as m, n, v,y, etc., and in the ovals of such letters as a, d, g, and q;also in the loops of loop letters and in the connectionsbetween letters within words. In the modern systemsof penmanship the rounded forms are taught and thisform is here considered as normal, giving two classes forthe factor.

What we term the eyed form, our third class, is illus-trated in Fig. 2, and is seen to be a small loop whichmay make its appearance in almost any small letter andis sometimes found in the capitals. This is a small thingto look for, but a big thing to find, being a persistentcharacteristic with many writers and likewise a verysignificant one. Being^a greater departure from normal,it is given first consideration in the classification ; in otherwords, look for the eyed_jorms first^ and if present,classify as No. 3 ; if not present, then base the class onthe angularity or rotundity.

9

7/29/2019 1922 - Handwriting

44/158

WuoHaMI

I

Xa

ft

7/29/2019 1922 - Handwriting

45/158

5o

>I

T5

fe .23 -a5 E

6en

a

U ^o

7/29/2019 1922 - Handwriting

54/158

CLASSIFICATION OF HANDWRITINGIn the case of a stiff, unyielding pen, it may occasionallybe necessary to resort to some magnification to obtaina true concept of the difference in pressure as evidencedby the depth of the nib impressions in the paper.

In the case of pencil writing the line of demarkationbetween the lightest and the heaviest strokes will not beso well defined but should, nevertheless, be sufficientlyevident when magnified to permit of their ready differ-entiation. See Fig. 5.

Rule IV. Where the difference in width between thelightest and heaviest strokes is imperceptible or veryslight, classify as No. 1.

Where the difference in width between the lightestand heaviest strokes is great, classify as No. 3.

Where the specimen does not fall clearly into classNo. 1 or class No. 3, classify as No. 2.

V. MovementOsborn says of movement:In the finger movement the letters are made

almost entirely by the action of the thumb, index,and middle fingers, the actual motion involving thesecond, and to a slight degree the third joints. Thismovement gives but little freedom of any kind,and especially but very slight lateral freedom. Itshows lack of clear-cut, smooth strokes, and con-tains numerous broad curves marked by somewhatirregular connections between letters and parts of

20

7/29/2019 1922 - Handwriting

55/158

7/29/2019 1922 - Handwriting

56/158

CLASSIFICATION OF HANDWRITINGletters, and is usually slow and labored. The handmovement is produced in most part by the actionof the hand as a whole, with the wrist as the centerof action, with but slight action of the fingers.What is known as the forearm or muscular move-ment really comes from the shoulder, but is themovement of the hand and arm, with the arm sup-ported on the muscular portion of the forearm withthe elbow as the center of lateral motion. It ispossible to Avrite entirely with the forearm move-ment without any separate action of the hand andringers, and many superior penmen write in thismanner; but the easiest, most rapid, and most per-fect writing is that produced with the forearmmovement Used in connection with a slight actionof the hand and fingers by which the small parts ofthe writing are produced. In the forearm move-ment, the stroke itself shows speed, force, freedom,continuity of motion, uniformity of pressure, andusually a uniform base line. The whole arm move-ment is the action of the entire arm without rest.In the forearm movement it is obvious that themotion and its source are nearly the same as in thewhole arm movement, but the forearm rests on asupport of desk or table. This rest restricts some-what the extent of the movement, but with suchsupport the movement is under somewhat better

. control. The whole arm movement is employed inornamental penmanship, in blackboard writing,and by a few writers in forming all the capitals.

22

7/29/2019 1922 - Handwriting

57/158

CLASSIFICATIONConsistent with the fundamentals of our system, we

have adopted the finger movements as one extreme andthe forearm movement as the other; between these twoextremes is what we call the compound movement,which includes the hand (or wrist movement as it issometimes called) and any combination of the threemovements. The whole arm movement is not oftenused in the ordinary writing, except perhaps with thewriter standing and without arm support or under otherunusual conditions; and as there is no Avell defined lineof demarkation between the writing produced by thewhole arm and by the forearm movements, they are bothclassed here under forearm or No. 3.

The finger movement is the one first employed inchildhood ; it is generally used by illiterates and by thoseunskilled in the art of penmanship. Most of the newvertical writing is executed with this movement. Fingerwriting appears formal, slow and labored; is usuallyshaded and without dash or flourish, having but littlelateral freedom; shows lack of clear-cut, smooth strokes;often shows irregular connections and uneven spacingbetween the letters and uneven alignment with the writ-ing line; sometimes contains broad curves and ovals;frequently evidences decreasing size of letters due to therestricted motion between the shifts of the hand neces-sary to carry the writing to the end of the line.

On the other hand, writing executed by the forearmmovement is conspicuous for its speed, force, freedom

23

7/29/2019 1922 - Handwriting

58/158

CLASSIFICATION OF HANDWRITINGof strokes, continuity of motion, uniformity of pressurewhich manifests itself in lack of shading, except, per-haps, on the terminals; it usually shows a uniform baseline. While it is true that the most perfect writing isproduced by this movement, it is also true that wherethe writer lacks natural or acquired facility in penman-ship, the forearm movement will result in a greater de-parture from copy book standards than will the fingermovement. Thus, were the first writing efforts of thechild executed by the forearm movement instead of thefinger movement, legibility would be much harder toattain.

To one who writes with the forearm movement orwho has practiced the exercises used in teaching themovement, there should be no difficulty in differentiat-ing it from the finger movement if one but recalls theeasy swinging motion imparted to the hand by the move-ment of the arm with the under muscles of the fore-arm resting on the desk. A little forethought and somepractice will render the classification of this factorsimple. Fig. 6 shows several specimens of each move-ment.

Rule V.If the specimen appears slow and labored,shows lack of freedom and clear-cut, smooth strokes,contains numerous broad curves, with irregular connec-tions, heavy shading, and decreasing size, classify asNo. 1.

24

7/29/2019 1922 - Handwriting

59/158

aJ o

-ocp=3 Bij o

7/29/2019 1922 - Handwriting

60/158

CLASSIFICATION OF HANDWRITINGIf it does not appear slow and labored, shows some

freedom and medium shading, classify as No. 2.If it shows speed, force, freedom, continuity of

motion, uniformity of pressure (that is, little or no study-ing), classify as No. 3.

VI. EmbellishmentThe older systems of handwriting show a muchgreater degree of embellishment than the modern ones;

in fact, some of the copy books of the modern verticalsystem appear to aim at extreme simplicity of form,both as to capitals and small letters. We are here con-cerned only with those flourishes, grace lines and super-fluous strokes useful for ornamentation exclusively ; thatis, strokes not essential to legibility. Fig. 7 shouldconvey the idea clearly as to what is here meant asEmbellishment.

Rule VI.If flourishes or superfluous strokesappear in capitals or small letters in sufficient numberto indicate they are habitual, classify as No. 3.

If no such strokes are found, classify as No. 1.If such strokes appear only occasionally, classify as

No. 2.VII. Terminals

We are here concerned with the general inclinationof the terminal or finishing stroke of both capitals andsmall letters. Generally speaking, the terminals of the

26

7/29/2019 1922 - Handwriting

61/158

7/29/2019 1922 - Handwriting

62/158

CLASSIFICATION OF HANDWRITINGsmall letters give a truer index to the habit of the writerthan do those of the capitals, the latter showing a greatervariation in the majority of writings than the small let-ter terminals. Dropped terminals, that is, finishingstrokes not completed as found frequently in letter s,occasionally in letter o and in the lower loop letters f,g, j, y, and z, are not considered here.

With a sufficient amount of material this factorpresents no difficulty but it may occasionally happenthat a single signature may lack an ample number ofterminal strokes to definitely determine the habit of thewriter, some terminals having one inclination and someanother. Preponderance of the one or the other is theguide in such cases. If necessary recourse may herebe had to a reference number (as is done in fingerprintclassification in doubtful cases) by first noting the classnumber that appears most applicable, and then notingabove or below this the next class number, and search-ing the specimen both ways. See Fig. 8.

Rule VII.If the inclination of the terminal strokesis upward, classify as No. 1.If horizontal, classify as No. 2.If downward, classify as No. 3.

VIII. SlantThe degree of slant above horizontal, or the base

line, is measured with a transparent protractor, as illus-28

7/29/2019 1922 - Handwriting

63/158

7/29/2019 1922 - Handwriting

64/158

CLASSIFICATION OF HANDWRITINGtrated in Fig. 9. In making these measurements besure that the horizontal line of the protractor liesparallel with the writing line and then note the slant ofthe various letters. In loop letters such as b, f, h, k,and 1, the axis of the loop is measured rather than theupward or downward strokes. The measurements of afew, preferably double-space, letters will suffice to givethe average slant of the writing.

Slant is a characteristic that becomes quite signifi-cant under certain conditions and with many writersis one of the most fixed of habits. Writing ranges inslant all the way from 35 degrees above horizontal tothe right to 50 or more degrees to the left of vertical.The limits set for classification purposes here shouldgive a fairly even distribution in the files for all systemsof writing in vogue at present.

Rule VIIIIf the average slant measures less than60 degrees, classify as No. 1.

If between 60 and 80 degree's inclusive, classify asNo. 2.

If greater than 80 degrees, classify as No. 3.Where there are a considerable number of specimens

to be classified at one time, it is our experience that theoperation is greatly facilitated by examining all of themfrom the viewpoint of one factor only at a time ; that isto say, they are all studied first as to Form, and the 1,2, or 3 designating the class of that factor noted oneach; next they are studied as to Skill, and the 1, 2, or

30

7/29/2019 1922 - Handwriting

65/158

Class 1.Less than 60 degrees. Note that the 50-degree line is parallel to theaxes of "11."

Class 2.60 to 80 degrees. Note that the 70-degree line subtends the axis of"1" in "Police."

PS- 100"= =Class 3.More than 80 degrees. The average slant is about 100varies from 95 to over 110.

Fig. 9.Illustrating the three classes of (VIII) Slant.It

7/29/2019 1922 - Handwriting

66/158

3.ssoO

ooI

I

H

7/29/2019 1922 - Handwriting

67/158

CLASSIFICATION3 designating the class noted on each, and so on untilea^hrSPeeiijmieBWbears a fis-ure consistiner of eisht djeritg.-if IS IS V V> Vii Vui

11 Z-

7/29/2019 1922 - Handwriting

68/158

H|eo

7/29/2019 1922 - Handwriting

69/158

CLASSIFICATION3 designating the class noted on each, and so on untileach specimen bears a figure consisting of eight digits.The eight factors here utilized, each divided intothree classes, will give 9,840 divisions in our file, whichshould be adequate for a collection of 100,000 speci-mens, since with an even distribution of this numberin the files there would be approximately only ten ineach group. However, other factors may be addedfrom time to time as required for larger collections andfor this purpose the additional factors mentioned in thetable on page 32 are suggested.

Thus we find that by using these additional fourfactors we would have 1,190,640 divisions in our file,which would more than care for a collection of 10,000,-000 specimens.

7/29/2019 1922 - Handwriting

70/158

CHAPTER IVINDEXING AND FILING

It will occasionally become necessary to take fromthe files the writing specimen of a particular knownauthor for comparison with writings of unknown originor for other purposes, and to enable one to locate in thefiles the desired specimen, every specimen of knownorigin is indexed alphabetically by the name or namesused by the writer. This is done on a 5 x 3 inch cardafter the specimen has been classified and made readyfor filing. These indices should bear, near the top ofthe card, the name of the writer, his criminal specialty,the police case number and his record number respec-tively; and near the bottom, the handwriting classifica-tion number. It is suggested that this classificationnumber be preceded by the initials "H. W." (handwrit-ing) to prevent the possibility of these indices beingconfused with others of a similar nature. These indicesare filed alphabetically.

An 8x5 inch vertical file is recommended for stor-ing the handwriting collection, as this is a standard sizein filing cabinets and will be found to answer the ordi-nary needs. Specimens consisting of a signature only,or of a name and address and other data, may be writ-

34

7/29/2019 1922 - Handwriting

71/158

INDEXING AND FILINGten or pasted on an 8 x 5 specimen card, as suggestedin Chapter I. But where the specimen consists of aconsiderable amount of writing, it may be convenientlyenclosed in an 8 x 5 folder, with a sample of the writ-ing on the front of the folder, thus frequently avoidingthe need for opening it and unnecessarily handling thecontents. The classification number, and the other datasuggested in Chapter I, should also be noted on thefront of the folder. Most written documents by fold-ing lengthwise in their original condition, or by trim-ming down to eight inches wide, may be convenientlyenclosed in an 8 x 5 folder.

At the outset the collection may be stored in a singledrawer and filed numerically without guide cards. Asthe collection increases in size, it will, of course, be neces-sary to install guide cards for convenience in searchingand filing but a complete set of guides will not be neededuntil the collection has grown to considerable propor-tions. A complete set would be numbered as follows

11,111,111 11,111,112 11,111,11311,111,121 11,111,122 11,111,12311,111,131 11,111,132 11,111,13311,111,211 11,111,212 11,111,21311,111,221 11,111,222 11,111,22311,111,231 11,111,232 11,111,23311,111,311 11,111,312 11,111,31311,111,321 11,111,322 11,111,32311,111.331 11,111,332 11 ?H1 ?333

35

7/29/2019 1922 - Handwriting

72/158

CLASSIFICATION OF HANDWRITINGAnd so on until the guide numbered 33,333,333 isreached. Three- or nine-cut guides may be used forthis purpose. If three-cut are used, they should benumbered in the order shown above, which would resultin the numbers on the first-cut guides ending in 1, thesecond-cut in 2, and the third-cut in 3.

Another simple scheme for numbering and arrang-ing the guides consists in using eight-cut guides, eachcut representing one of the eight factors. Of the first-cut three are used and numbered 1, 2, and 3, to rep-resent the three classes of the first factor. Behind eachof these three guides are placed three of the second-cut, of which each set of three is numbered as before,1, 2, and 3, representing the three classes of the secondfactor. Behind each of these nine are placed three ofthe third-cut guides, of which each set of three is alsonumbered 1, 2, and 3, representing the three classes ofthe third factor; and so on until we have used all ofthe eight cuts. The guides required to inaugurate thisscheme are shown in table on page 37.An ideal cabinet for use in connection with thislatter arrangement would consist of nine draAvers, 8x5,three rows of three drawers each, the lower portion ofthe cabinet adapted for filing the 5x3 alphabeticalindices, similar to the cabinets used for filing Bertillonrecords. The upper row of three drawers would allcontain specimens of Class 1 of Form; the middle, orsecond row, would all contain specimens of Class 2 of

36

7/29/2019 1922 - Handwriting

73/158

INDEXING AND FILINGPosition

7/29/2019 1922 - Handwriting

74/158

CLASSIFICATION OF HANDWRITINGNow supposing our specimen to be searched or filed

is numbered 11,111,111, we would first turn to theupper horizontal row of drawers; then to the first orleft-hand drawer of that row; then to the first or No. 1of the first-cut guides in that drawer; behind that wewould next select the first or No. 1 of the second-cutguides; behind that No. 1 of the third-cut guides andso on until we came to the group of cards filed behindthe last or sixth-cut guide. In this group would befound specimens agreeing in eight significant particu-lars with the writing in question, and all that remainsof our task is to determine whether any of these are theproduct of the same hand. Bearing in mind that wehave nearly ten thousand divisions, even in so large acollection as one hundred thousand specimens, assumingthey are evenly distributed throughout the files, actualcomparison would have to be made with only ten ineach group.

Our search in each group may be further facilitatedby the use of colored specimen cards to segregate thespecimens according to the race and sex when known.For this purpose specimen cards containing colors men-tioned in list on page 39 may be used.

If the color scheme is not utilized, the same resultsmay be obtained by noting the race and sex of thewriter in the upper right-hand corner of the card andsegregating the cards in each group in accordance there-with, See Fig. 36. Occasionally anonymous writ-

38

7/29/2019 1922 - Handwriting

75/158

INDEXING AND FILINGings will bear a sufficiently conclusive indication of aforeign hand to justify their being filed on a "foreign"card.

Card Color Race and SexWhite White maleGreen White foreignersPink All femalesChocolate Black maleRed Indian or Mexican maleYellow Japanese or Chinese maleBrown Brown (Malay) male

Search may be further facilitated by the simple ex-pedient of noting at the top of the specimen card thehair-color, eye-color, height, weight and age whenknown, and segregating the cards of each color in eachgroup according to height and age in increasing order.Thus, supposing our questioned document were knownto have been written by a negro of six feet or more, wewould turn to the group corresponding to the classifica-tion number, select only the chocolate-colored cards andmake comparison only with those bearing a height offive feet and ten inches or more and of the approximateage of the unknown writer. The hair and eye color willfrequently prove useful in eliminating some of the speci-mens bearing descriptions otherwise approximatingthat on the card being searched, and the weight mayoccasionally be useful for the same purpose.

39

7/29/2019 1922 - Handwriting

76/158

CHAPTER VIDENTIFICATION

We have learned something of the theory of classifi-cation and filing and we now come to what is perhapsthe most difficult phase of our problemthat of identifi-cation of handwriting. With our specimen properly-classified, we proceed to a definite group of specimensin the file, all of which agree precisely in eight particu-lars with the one in question, but this concurrence ofidentifying characteristics does not, of course, establishor constitute identity. So we are here confronted withthe question, "To what extent must the characteristicsof two writings coincide to positively establish theiridentity?"

Osborn states:Two writings, one of which is disputed, may

appear very different in general appearance, butmay contain so many small but peculiar and per-sistent characteristics that the conclusion of identityis irresistible; and again two writings may be verysimilar in certain general features, but may differin so many significant but inconspicuous particu-lars as to show that they were undoubtedly written

40

7/29/2019 1922 - Handwriting

77/158

IDENTIFICATIONby two different writers. When inconspicuouscharacteristics persistently and consistently diverge,the conclusion must be reached that two such writ-ings are not by the same hand ; when they coincidein sufficient number the conclusion is reached thatthey were by the same writer.In another chapter Osborn further states in this con-

nection :The principle underlying the identification of a

handwriting is the same as that by which anythingwith a great many possible variations is identifiedas belonging to that class or being that particularthing. It is first necessary to establish the stand-ard, and then identity or difference is shown by acareful comparison of all elements, features orcharacteristics which altogether constitute the basisfor a conclusion. The force of the conclusion isnaturally governed by the number and significanceof the points under consideration, ranging from amere conjecture up to what amounts to moral cer-tainty.

There are many close analogies between theidentification of an individual by bodily character-istics and the identification of a handwriting byexamination of its various elements. As we know,in some instances identification is practically cer-tain, but in others only probable, or there may beinsufficient basis for any opinion. If an individualis sought for who is definitely described as follows

41

7/29/2019 1922 - Handwriting

78/158

CLASSIFICATION OF HANDWRITING(1) exactly five feet, eleven and one-fourth inchesin height, (2) blue eyes, (3) brown hair, and inaddition has (4) lost his left thumb, and the (5)lower part of his right ear, and has (6) a moleon his left temple one-half inch in diameter, and(7) a tattooed anchor on the back of his left hand,and (8) a narrow scar five inches long on his rightforearm, and an individual is found who exactlymatches these eight points of identity, we say with-out hesitation that we have undoubtedly found theman described. We promptly conclude on theseeight points alone .that this man differs from allthe other millions of men on the earth. Theoret-ically there may be other such men, but we sayconfidently that it is so strongly against commonsense and all experience to expect it that it is prac-tically impossible.

This problem is capable of a mathematical solu-tion if we first agree on the basis for the calculation.It is possible to show mathematically how vastlyimprobable it would be for these eight separatepoints of identity to coincide in two individuals.We must first determine how often, or rather howseldom, each feature will be found separately, andthen by a mathematical formula as fixed as themultiplication table, we determine how often coin-cidence of all the features may be expected. Thisformula, as given by Professor Simon Newcomb,is as follows: "The probability of concurrence ofall the events is equal to the continued product of

42,

7/29/2019 1922 - Handwriting

79/158

IDENTIFICATIONthe probabilities of all the separate events." Ifone thing will occur once in twenty times and an-other once in twenty times, the probability of thetwo occurring in conjunction is represented by thefraction which is the product of one-twentieth andone-twentieth, or one four-hundredth.

For the purpose of showing the practical impos-sibility of these eight personal features all beingexactly duplicated in two individuals we can makeso small the fraction representing how frequentlyeach point may be found that it is certainly withinthe fact and will be promptly granted. For num-ber one, the exact height, we will say that in everyten men we shall find one who exactly measures astherein described; of the second or eye color, onein three; of the third, one in four; of the remain-ing five points, we will say that out of every twohundred men, one will be found precisely answer-ing one of each of the peculiar descriptions speci-fied, although for some of these accidental andunusual features more than one in ten thousandwould certainly be too many to expect.We now have our separate events representedby eight fractions, one-third, one-fourth, one-tenth,and five fractions each of one two-hundredth.Now, assuming that our problem is to determinehow frequently all these peculiarities and

ele-ments that have accidentally combined in this oneindividual would be likely to be exactly du-plicated in another individual, we find by apply-

43

7/29/2019 1922 - Handwriting

80/158

CLASSIFICATION OF HANDWRITINGing Professor Newcomb's rule that the continuedproduct or the mathematical probability is one in38,400,000,000,000, or more than thirty thousandtimes the population of the globe, and we naturallyconclude if we find an individual who answers thedescription, that we have found the man, and thatthere is not another man who has had the samethings happen to him.

Let us see just what a proposition it is to expectcomplete identity in two handwritings. We mustimagine groups of the variations of each of theeighty or more characters in English script withfrom a dozen to a hundred or more variations ineach group. Then we take one variation of eachof the eighty characters until we have a completeset. The proposition now is to go out and find awriter who will duplicate in his writing every oneof these selected characteristics. The improbabilityis simply ridiculous and we say at once it cannotbe done. Even if the probability of making a par-ticular form in each group were as low as one-fifth,to have this combined on the same ratio with a par-ticular form in all the other groups would make aprobability represented by a fraction with one forthe numerator and the eightieth power of five as adenominator.

We may conclude from the foregoing that in anygiven case we should be able to state with some degreeof accuracy the chances of a questioned writing being

44

7/29/2019 1922 - Handwriting

81/158

IDENTIFICATIONidentical with or different from another writing by usingas a basis for our computation the very conservativefigure of one-fifth (that is to say, one chance in five ofthe same particular Avriting characteristic being pro-duced habitually by two different writers) as suggestedby Osborn.

As a practical application of this theory, let us sup-pose our specimen in question consists of a name andaddress in which the letter "a" occurs, say, six times,and five of them are found to possess some inconspicu-ous identifying characteristic, such as an eye at the endof the oval. We set this down as one habitual form.Next we find the Greek form of letter "e" with the tophalf larger than the lower half in sufficient number toprove it is habitual with this writer; we note this as asecond habitual form. Looking further, we discoverseveral "f's" with the lower loop about twice the sizeof the upper loop and we note this as a third habitualform. And so on until we have found, say, eight char-acteristic forms of this nature in our questioned writing.In the group of specimens bearing the same classifica-tion number, we may find one that arrests our attentionby bearing some resemblance to the questioned writingand we proceed to check this over. We discover thatwhile the pictorial aspect has been somewhat alteredby the adoption of a backslant and heavier pressure,together with a change in the more conspicuous featuresof the capitals, there is a concurrence in the eight par-

45

7/29/2019 1922 - Handwriting

82/158

CLASSIFICATION OF HANDWRITINGticulars noted in the questioned writing and we may-state definitely that the chances of the two writingshaving been produced by two different individuals isone in 390,625, or five raised to the eighth power.

But we must bear in mind in this connection that allgenuine writing shows some natural variationthat is,the same forms and characteristics will not invariablybe usedbut this variation is usually manifested onlyin the more superficial parts. An identifying character-istic, therefore, to be of value, must be proven to behabitual with a particular writer and not merely occa-sional or exceptional; but since the writing process asa whole is one of the most permanent and unconsciousof human habits, we should be able to establish a con-vincing number of these habitual traits even in a smallamount of writing. In the above example the eightidentifying characteristics are assumed to be' habitual.Now supposing that instead of the differences enu-merated (pictorial aspect, slant, pressure, and capitals)we had found a divergence in a number of inconspicu-ous features (even two or three would suffice in thiscase) and these divergencies persisted throughout, ourestimate of the probabilities of identity would bematerially changed. Therefore, we must look for dif-ferences of form and other features as well as similari-ties, before we can determine the chances of identityor non-identity of two writings.

However, in actual practice we will find that the46

7/29/2019 1922 - Handwriting

83/158

IDENTIFICATIONprocedure outlined above will seldom be necessary forthe reason that once we have become conversant withhandwriting analysis, we will know more or less in-tuitively when two writings are the product of the samehand, especially in those cases where only an unskillfulattempt at disguise has been made which, fortunately,are largely in the majority. But this faculty of directcognition must be developed by systematic study beforeit can be relied upon implicitly, and this developmentalprocess will best be enhanced by a consideration of allthe elements that go to make up individuality in hand-writing.

1. Alignment of Writing.This is the relation of theletters of a word or a line of writing to an actual orimaginary base line and it is largely dependent uponthe movement, manner of holding the pen and the designof letters. With the elbow acting as the pivot and thearm perpendicular to the writing line, the alignmentis often perfectly straight across the page even onunruled paper. With the wrist as the center of motion,the lines may be made up of short arcs, representingthe reach of the hand with the wrist at rest when movedaround to the right as far as the hand will reach. Themost uneven alignment results when the arm is toofar around to the right or the paper too far to the leftso that the lateral motions of the hand to the right, ex-tend above the base line of the writing. Certain lettersare frequently made too high or too low by some writers.

47

7/29/2019 1922 - Handwriting

84/158

X TheSrstKationalBank WI

or uRUELir ^

^.w^vAn ^S^/v^- ^^-r^==t==r^^_Dooms_J^,^.._^Avyi&a^

TheI^T^pALBank JgBOTIffilEf^r0^1A._5^*iA~JJ-^M_^A'9-^ Ho. 9-5)A

BsnOTHEOBDERO^M-^l^> >J>$&Sa2^,*W^^"^a_3u. ^- - ^^-~^_ ??rST7v& Dollars

_ . JXb.-_r^^...^

7/29/2019 1922 - Handwriting

85/158

Praddait Statm FaLrraticm, al Lvbtm. QPreaident.

Fig. 10B.Illustrating faulty Alisnmerit.The first signature has a downward inclination. Classification No. 82,122,121.Both names in the second signature have a downward inclination. Classifica-

tion No. 33,123,221.In the third there is a slight upward trend. Classification No. 32,122,321.

There is a slight upward trend in the fourth. Classification No. 22,121,231.1

The bottom signature if convexed upward. Classification No. 31,122, 3

7/29/2019 1922 - Handwriting

86/158

Fig. 11AAngles, (a) Illustrates the uniformity of Angle Values found in copy-book style and manner of drawing the lines.

Pig. llBAngles. (6) Mr. Lindley claimed the signature shown above was aforgery. The Angle Value test alone proved this a forgery. Each of theselines should be compared with the preceding line, with which it may beparallel or form an angle at top or bottom.

7/29/2019 1922 - Handwriting

87/158

Fig. 11C.Angles, (c) The upper is a natural signature. The middle, by thesame writer, is effectively disguised as to Alignn ent, Form, Line Quality,Movement, Pen Position, Proportion, Skill, Slant, and Spacing, but thedotted line is the only one of the thirteen that departs radically from theAngle Values of the natural writing. The lower is a simulation of the upperby another writer and here quite a different picture is presented.

7/29/2019 1922 - Handwriting

88/158

Berkeley, Cal. _Amtk^> -^ i-*- >i> 2frt:y^.Ll:y-\*jat,-IMa:l:^V>frty

Hcikcky Ifcmk ol" Savings aiulTmsl (loiniKiiiV'*" ~ " ~lH5T'NATIOfgAL I^NK O rgQERK ET LEY Jl'-\>^"0'JIII. OlilttKIl

f Q.3

c^w-t

"nr.HM/j.la.avi.. -AJj^Ll^ 193_L_ Mo 3^.ASSOClATtD WtTn |R5T#ATiONAL#BAK>N d *E PKE L EY t *'

0-5,

**> * * rt'ffHg^*"*?,?"* 8 * -~ fe^v- tmrAYTHTIIEllKmilHiK 1 ^^V'

Fig. 11D.Angles, {d) Mr. Moore claimed the middle check to be a forgery andsubmitted a number of his canceled checks for comparison, of which theupper and lower were selected at random for the Angle Value test. Tracingpaper was placed over the checks and the axis lines drawn thereon. Checkswere photographed, then the tracing paper removed and this latter photo-graphed separately as shown in the reduced illustration. Note that in thetwo standard signatures lines 20 and 21 are the only ones between whichthere is not complete coincidence, whereas in the questioned signature lines1, 2, 3, 14, 15, 16, and 21 depart more or less from the Angle Values found inthe standard writing.

7/29/2019 1922 - Handwriting

89/158

Fig. HE.Angles. See especially 11D.

7/29/2019 1922 - Handwriting

90/158

CLASSIFICATION OF HANDWRITINGThe alignment may be straight but with an upward

or downward tendency; it may be convex or concave orirregular or sinuous. See Fig. 10.

2. AnglesDr. Locard, Director of the PoliceLaboratory of Lyons, France, states:

If a person imitates the slope of a handwritinghe preserves, in spite of himself, the proportionsof the angle values peculiar to his own hand. Theresult is that as the average axes of the differentletters are not parallel their prolongations willmeet; and by assembling all the points of inter-section a characteristic design will be obtained.

In order to apply the angle value test effectively,it is best to select from the two writings being examined,the same words, or words having the same letters in asnearly the same sequence as possible. For the purposeof determining and illustrating angle values, the speci-mens under consideration are placed on a table or boardand covered with a piece of tracing paper of sufficientsize, which is fastened down firmly with thumb tacksor. by other means. Lines are then drawn on the trac-ing paper over the straight lines of the letters or theiraxes as shown in Fig. 11. This is a very effective testfor forgery.

3. Arrangement.The arrangement of the writingon the page is often characteristic with reference toindentations, paragraphing, spacing at top and bottom,

54

7/29/2019 1922 - Handwriting

91/158

IDENTIFICATIONdate line, signature, etc. This element is of especialimportance in the analysis of questioned checks, whereinthe following points should be noted:

(a) Manner of writing the date; whether month isexpressed by numerals or spelled out in full or abbre-viated ; also manner of writing check number.

(b) Distance of the beginning of the writing fromthe printed headings.

(c) Manner of writing the amount in figures andin words, and the various positions of the word "and"and the sign "&" with its variations.

(d) The varying forms of dash line between theamount in words and the printed heading "dollars."

(e) The manner of denoting cents.Fig. 12 illustrates several characteristic arrange-

ments of the writing on checks.4. Averages and Divergencies.In an article de-

scribing methods devised and used by Dr. Locard,Director of the Police Laboratory in Lyons, France,Mr. E. Weiss states:

For a given subject, his handwriting pos-sesses, in spite of himself, in spite of all dis-guises, a certain number of constant peculiaritiesthat may be measured and that furnish tothe expert indications that are unmistakable.By means of this method, which is calledgraphometry, we compare the measurements of theheights of the different letters and their propor-

55

7/29/2019 1922 - Handwriting

92/158

P Ofo/t>J&r/*frHiC0.CLEARING HOUSC N9 16.l^J *, /r*t &A&'k

>M) -

DOLLAHS^^'Mi^r

Fig. 12A.Arrangement.(a) Note characteristic dash after name of payee and after amount in words;

also double curve to horizontal line for fractions. Classification No. 12,222,222.(c) Unusual arrangement. Classification No. 22,122,212.

7/29/2019 1922 - Handwriting

93/158

SanJosk.,CAi,.i?i^j '7 inig. n H--('en'ti{.\i,N"atioxaij1Jank. w^~= =-= f3-- b STATr.S IIKI'OSITAHV

(HI (MRIJIliD Bmic OrSflMMtaM

Fig. 12B.Arrangement.(6) Note proximity of writing to printed headings; long, sinuous dashes and

location of fractions as well as flourish line under signature. Classification No.22 221 113 for body of check, No. 22,122,331 for signature.

(d) Also unusual. Classification No. 32,122,133.

7/29/2019 1922 - Handwriting

94/158

CLASSIFICATION OF HANDWRITINGtions to each other. For a given letter, the averageheight is constant relative to that of the other let-ters, even if the general scale of the writing bechanged. One who habitually makes large "s's"and small "i's" will maintain these same propor-tions.

In difficult cases it may occasionally be necessary toresort to this test of averages and divergencies there-from in order to form a definite conclusion as to identityor nonidentity between two writings. The procedureis this

Measure the height of all the single-space letterswith a finely divided rule, say 50ths of an inch, and findthe average thereof. Proceed in a similar manner withthe double- and treble-space letters.Next find the average height of each individual

single-, double- and treble-space letter.Suppose the average height for all the single-space

letters is -^ and we note that the "a's" average -$, thatis, -^ above the general average; and the "e's" average^q, or g2o below the general average for single-spaceletters. Next we find that the double-space lettersaverage ^ and that the "h's" average ^%, or -^ abovethe general average; and that the "Fs" average ^, or52o below the general average; and so on with the treble-space letters.Next we proceed in a similar manner with the seconddocument in question, in which the writing may be

58

7/29/2019 1922 - Handwriting

95/158

IDENTIFICATIONlarger or smaller than in the first, but this fact will notalter the results. After finding the general averages ofthe three size letters and the averages of the individualletters in each group, we proceed to note which of themdiverge from the general average and whether above orbelow. Thus we may find that the "a's" are below thegeneral average for the single-space letters instead ofabove, as in the first document; that the "e's" are abovethe general average; similarly that the "h's" are below,and that the "l's" possibly do not diverge whatever fromthe general average. We may now summarize our find-ings as follows:

7/29/2019 1922 - Handwriting

96/158

CLASSIFICATION OF HANDWRITINGthe questioned and the standard writing in the sameway. If it is found that the divergencies are approxi-mately equal and in the same direction, that is, aboveor below, this is a very strong indication of identity ; butif in opposite directions, as shown above, it is positive

50^ CU tWn 9/o ^Mim-urryidb-iii/lit "tfHf( fa2.0i.11111Jl5V*131%U10

16-5A3%i-

wPig. 13.Averages and Divergencies. Solid line represents the standard writing;

dotted line the questioned. In each case the straight lines represent thegeneral average for letters of the same height and the curves show divergenciestherefrom of individual letters.

proof of nonidentity if based on a sufficient amount ofmaterial. See Fig. 13.

5. Characteristics.The term characteristics in itsapplication to handwriting embraces all those elementsthat go to make up individuality but what we are hereconcerned with under this heading are those originalform characteristics which differ or depart from copy-

60

7/29/2019 1922 - Handwriting

97/158

IDENTIFICATIONbook standards, called "trade-marks" in the parlance ofthe expert, and which are frequently met with in checksand other writings. Strange to say we have seen someof these persist throughout a series of checks in whichthere was a strenuous attempt at disguise made in otherparticulars in an effort to lead the authorities to believethat the checks were not all the work of one man. InFig. 14 are illustrated a few of these trade-marks.

6. CurvesWith the majority of forearm writersand with some finger writers the degree of curvaturefound in parts of certain capitals is a somewhat constantand therefore an important factor. A simple methodfor testing this element is illustrated in Fig. 15. Thedocument is placed under a sheet of tracing paper, asexplained for (2) Angle Values, one of the points ofa draughtsman's compass is placed at the center of thearc and the projection of the arc is represented by lightor dotted lines drawn with the other point of the com-pass. Parallelisms or differences of degree of curva-ture between two writings are graphically shown in thisway.

7. Form The particular form of each letter in thetwo writings in question should be studied minutely,each document separately, individual characteristicstabulated and then the two compared with each other.It should be borne in mind, however, that even frequentcoincidence in the general form of letters in two writ-ings does not establish their identity where there is

61

7/29/2019 1922 - Handwriting

98/158

0i '/''-' 19 -v- |N \l.ll V\K - - i

^/ ibi/ nT^T'PaffiS HftBtllMftQBSMi BSfflBB

fcA $*tf^pa^kfK^X^?. /& i9i/ N? '/

Fig. 14A.Characteristics, (a) Series of checks all by the same writer andNote date always expressed wholly by figures, " % " before the amount

7/29/2019 1922 - Handwriting

99/158

T?%4ftrs|Sajfaa$IBank *

r

I(the 0mt Saghnud feitk^ J

(HtrJD&t &tttmin 1 IBank

TLsr^ X 3 O/fe^ 22$*

* et^u- i/ci gCexthai.Nationai.Hank '

extending over a period of nearly three years, yet all "trade-marked" similarly.and characteristic form of "no/100." Classification No. 23,223,211.

7/29/2019 1922 - Handwriting

100/158

CLASSIFICATION OF HANDWRITINGmarked dissimilarity in other important features, asthere must necessarily be a considerable agreement ofform displayed by two different writers of the samesystem of handwriting. Here, as with other factors,

(Thrifii&t&tf10u

(Eltrlfiisi&ttunuil I'mi k

d&7^*" I)0LIARS

4^Fig 14B(6) Note trade-mark "XX" for expressing cents. ClassificationNo. 31,232,311.the small inconspicuous characteristics are the mostimportant in individualizing handwriting.

(a) Small "a" should be examined as to the formof its initial stroke, if present; the form of the oval,whether circular or elongated and, if elongated, thedirection of its axis; whether open or closed at the top

64

7/29/2019 1922 - Handwriting

101/158

IDENTIFICATIONand whether the first stroke of the oval is higher orlower than the last; and the form of its connectingstroke with the succeeding letter.

(6) Examine "b" as to the height at which the upstroke and the down stroke of the loop cross ; the widthof the loop and particularly the shape of the top of theloop, whether broad and rounded or flat, or whether

Fig. 15.Curves. Note that degree of curvature of first "A" is smaller thanthe other two and second slightly smaller than the third. Also note thatposition of the circle relative to the arc it subtends is different in each case.Similar differences are to be found in the "B's."

angular or irregular ; whether the first down stroke goesclear to the writing line; height above writing line atwhich the horizontal bar is made and the shape of thebar, which may be straight or may be sufficiently. curvedto give the "b" the appearance of "li."

(c) Letter "c" may be begun without an initial strokeand it has several distinct forms; it may start with asmall eyelet, a mere dot or a tick; when connected with

65

7/29/2019 1922 - Handwriting

102/158

CLASSIFICATION OF HANDWRITINGthe preceding letter, the link may make the/'c" intoan "e."

(d) Examine letter "d" relative to the oval thesame as "a"; note also whether the staff (the partfollowing the oval) is looped and at what point relativeto the oval the down stroke of the staff separates fromthe up stroke. The terminal stroke of final "d" is oftendropped.

(e) Note the width of the eyelet and the shape ofthe top, whether broad or narrow, whether rounded orangular. Very narrow "e's" often have the appearanceof "i." The Greek form of "e" shows considerablevariation; the top half may be larger or smaller thanthe lower half; between the two halves there may occuran eyelet; the lower half may be a complete circle.

(/) Examine letter "f" as to the relative size ofthe two loops; as to the point at which the up and thedown strokes of the two loops cross; as to the shape ofthe loops, the same as for letter "b"; and the shape ofthe connecting link with the following letter.

(g) Examine letter "g" relative to the oval the sameas for letter "a"; and relative to the loop the same asfor "f."

(h) Examine letter "h" relative to the loop thesame as for letter "b"; note the shape of the arc form-ing the latter half of the letter, whether rounded orangular; how far it is retraced back up the staff and

66

7/29/2019 1922 - Handwriting

103/158

IDENTIFICATIONthe form of the connecting link with the followingletter.

(i) Letter "i" is sometimes made in the form of"e"; note how far the down stroke retraces the upstroke; note position of the dot as to its height abovethe letter and whether to right or left ; note particularlythe form of the dot, which may be wedge-shaped andsloping in any direction, a horizontal dash, a small circleor semicircle, a small "v," a perfect dot or comma-shaped with the tail in any direction.

(j) Examine the first half of the "j" the same asfor letter "i" and the lower loop the same as for let-ter "f."

(k) Examine the loop of letter "k" the same as for"b." The last part of this letter is known technicallyas a "buckle" and this has a diversity of forms, whichshould be carefully noted; note also relative height ofthe buckle to the loop.

(Z) Examine the loop of letter "1" the same asfor "b."

(m) Examine letter "m" as to the shape of the threearcs of which it is composed; as to their relative widthand height ; as to their relative separation and the extentof the retrace at the bottoms of the middle arc.

(n) Examine letter "n" the same as for "m." Whenthe arcs of this letter are angular it has the appearanceof a "u,"

67

7/29/2019 1922 - Handwriting

104/158

CLASSIFICATION OF HANDWRITING(o) Examine letter "o" relative to its oval the same

as letter "a"; and relative to the bar, the same as for"b." The bar, or link, may be carried down too far,giving it the appearance of letter "a."

(p) Note whether letter "p" has a loop at top orbottom; note point of separation of first up and downstrokes and the same for the down and up strokes be-low the line. The normal form for the last part of thisletter is the same as for letter "h" but it may take onthe form of an oval or circle, which should be carefullynoted as should also the form of the connecting link.

(q) The form of the oval of letter "q" should beexamined the same as for letter "a." The lower loopis sometimes written the same as in letter "g" but thenormal form is the reverse thereof, in which latter casethe point at which the last up stroke touches or crossesthe preceding stroke should be noted; also note formof the link.

(r) Letter "r" has several forms, the normal havingan angular top with a small tick formed by the firstup stroke passing slightly above the body of the letter;another form has an angular top without the tick; an-other has an eyelet in place of the tick ; and still anotheris rounded at the top, this latter form sometimes havingthe final stroke retracing the preceding stroke. Theangular top, with or without tick, may be nearly orquite square or the top part may be curved downwardto such an extent that it assumes the appearance of

- 68

7/29/2019 1922 - Handwriting

105/158

IDENTIFICATIONletter "u." The rounded top may be followed by ahorizontal bar, the same as letter "b."

(s) Examine letter "s" as to the form of its initialstroke; whether it has an eyelet at the top or a tick, oran eyelet at the bottom; the shape of the down strokeand the point at which the down stroke terminates andthe final stroke commences. This letter is frequentlymade in the form of the printed "s," especially at thebeginning of words; and at the end of words may haveits terminal dropped.

(t) Note whether letter "t" has a loop and at whatpoint the up and down strokes cross or part. The cross-ing of the "t" takes on a great variety of forms, whichshould be carefully noted ; they may be straight, sinuous,or curved; horizontal or inclined up or down; very lightor heavily shaded or shaded on one end or the other;long or short; or may be ticked on either end. Thecrossing may be placed entirely above the staff of theletter or to the right or left or may cross the staff atany point between the top and bottom; final "t" is oftencrossed by retracing the staff with the terminal strokeor by a small backward loop or buckle, with a horizontalterminal, or there may be no crossing whatever.

(u) Note whether the two tops of the letter "u"are even in height and whether they are eyed and notedistance between them; note form of the arc, whethercurved or angular.

(v) Note the form of the arc of the letter "v,"69

7/29/2019 1922 - Handwriting

106/158

CLASSIFICATION OF HANDWRITINGwhether angular or rounded, and the same of the arcat the bottom; note whether the beginning of thehorizontal bar forms are eyelet; note distance abovewriting line of the bar and whether bar is straight orcurvedit may be sufficiently curved to give the letterthe appearance of letter "u."

(ta) Apply same tests to letter "w" as to "u" and"v."

(x) Letter "x" has several variations of form, anyone of which may be quite characteristic. Note par-ticularly the crossing, if the crossed form is used, asto its length, shape, position, shading, and slant.

(y) Apply same tests