NEW GLACIAL FEATURES OF THE ·UPPER PALEOZOIC

ITARARE SUBGROUP IN THE STATE OF SAO PAULO,

BRAZIL

By

A. C. ROCHA CAMPOS (1), J. E. S.FARJALLAT (1) e R. YOSHIDA (1)

ABSTRACT

Elongated topographic forms composed of bed­rock veneered by conglomerate, diamictite andsandstone and linear concentration of clasts in­cluded within diamictite are described from theItarare Subgroup (Tubarao Group), in Sao PauloState, Southeastern Brazil. Another possible elon­gated molded structure composed of sediments onlyis also briefly.

The topographic structures are interpreted asstreamline forms molded by the flow of ice, andthe clast concentration appear to correspond to atype of boulder pavement.

These features constitute additional evidencefor a glacial mode of origin for the diamictitesand associated sediments of the Itarare Subgroup.

RESUMO

Corpos alongados compostos de embasamentocupeado POl' diamictito, co'nglomerado e urenito econcentracao hIorizontal de clastos incluidos emdiamictito sao descritos no Subgrupo Itarare (Neo­paleoz6ico), no Estado de Slio Paulo. Uma outraestrutura alongada cornposta somente de sedimen­tos c brevemente descrita.

Os corpos alongados slio interpretados como for­mas topograficas lineares moldadas pelo fluxo dasgcleiras neopaleozoicas e a ooncentracao de clastosparece corresponder a um tipo de pavimento declastos.

Essas feigoes constituem evidencias adicionaisa favor de uma origem glacial dos diamictitos erochas associadas do Subgrupo Itarare.

INTRODUCTION

During the last few years additionalevidence has been acumulating in favor of aglacial environment of origin for the rocksof the Itarare Subgroup (Rocha-Campos,1967a) which is a thick sequence of sediments

distributed along the eastern margin of theParana Basin.

Among the most important glacial fea­tures recently described, the occurrence ofstriated surfaces overlain by diamictites(Amaral, 1965; Bigarella et aI., 1967),channel-like bodies of sandstone and conglo­merate and wedge-bodies of sandstone inter­preted respectively as fossil eskers and«crevasse»-fillings open in frozen till (Frakeset al., 1968), confirm a glacial environmentduring deposition of the Itarare Subgroup.

Furthermore, these features have provi­ded new paleogeographic data for the deter­mination of ice flow direction during Gondwa­nic glaciation, that are consistent with thepaleogeographic picture already known (Mar­tin, 1961; Rocha-Campos, 1967a).

In the present paper elongated topogra­phic forms composed both of bedrock andsediments are interpreted as streamlinebodies molded by the flow of glacier ice, andhorizontal concentration of clasts similar tothe Pleistocene boulder pavements of thenorthern hemisphere are described from the

Itarare Subgroup, State of Sao Paulo. Refe­

rence is also made to another possible

elongated body composed of sediments only.

These features represent additional evidence

supporting a glacial' origin for the sediments,

and provide new data for the determination

of direction of glacier movement during the

upper Paleozoic glaciation.

(1) Departamento de Geologia e Paleontologia,Universidade de Sao Paulo, Slio Paulo, Brasil.

48 BOL. DA SOCIEDADE BRASILEIRA DE GEOLOGIA - V. 17. N " 1. 1968

Acknowledgments

The authors wish to thank to Carlos A.L. Isotta, Instructor at the Department ofGeology and Paleontology, to Miss MariaAngela Fornoni and Messrs. J . SebastianAraujo and Virginia Montezzo Neto, seniorstudents of Geology at the University of SaoPaulo, for their help during field work,Thanks are also extended to Dr. John F.Rigby, who read the manuscript and offeredvaluable suggestions,

GEOLOGICAL SITUATION

The features described occur in the cen­tral and northern part of the State of SaoPaulo, in the eastern outcrop belt of theTubarao Group, in Parana Basin, Brazil(fig. 1).

The rocks studied belong to the ItarareSubgroup, the lower unit of the TubaraoGroup in the State of Sao Paulo, accordingto the stratigraphic scheme presented by Ro­cha-Campos (1967a).

The stratigraphic position of the feature"is variable. At Mococa they occur at thelowermost part of the local section of theItarare Subgroup and include the crystallinebasement. At Jumirim the features studiedoccur at the upper part of the Itarare, appro­ximately 10-20 meters bel ow the contact withthe Tatui Formation (Roch a -Cam pos, 1967a),the uppermost unit of Tubarao Group in thearea.

The age of these rocks is not firmlyestablished yet. A lower Permian age wasrecently suggested for the upper half partof the Itarare Subgroup in central Sao PauloState (R ocha-Cam pos, 1967b ). Possibly thesame age can be extended to the lowersection.

DESCRIPTION OF THE FEATURESElongated topographic bodies

The features studied are exposed in bothfaces of a road cut roughly oriented towardSW-NE along the roadway between Casa

I••

a eo 1$0 Krrt'------>-' _ -'-_--J.

+'\+ !1\.+ \ .........·r ""'\.'-'"-:-.

+ -il. .J + + + + )+ ).-+ + + :1;.""""'-

+ + + + ++ + + + + +

Sorocoba+. sao PlIIJlp +\,;IHi 'ii;::ijiiHii ++ + + + + + + z:s'r+ + + + + ++ + + + +..J:.-..:t:. + +

'.,j. +(,' c==J Younger ro cks

H::{ {;:;:;::J Tubara o Group

[3 Basement

2000 Km100 0oI

1. L ocation of ou tc rops s tud ied in the State of Sao P au lo. Arrow s a t Mococa an d Sorocab a indicateaxes of elon gated bod ies . Point of a r row a t J u m ir im indica te ded uced di rection of Ice-movemen t.

A. C. ROCHA CAMPOS, J. E. S. FARJA LLAT e R. YOSHIDA - New glaciaL.. 49

Branca and Mococa, at kilometer 261,5, appro­ximately 3 km south of Mococa.

The local rocks are decomposed and werecovered by a crust of mud that was partiallyremoved to make observation easier.

The north and south faces of the roadcut are 20 meters apart from each other.

Correlation between the two faces allowed thereconstruction of two parallel, elongated bo­dies, designated A and B, which are semi­elliptic in section and 22 meters apart. Theaxis of structure A is elongated towardN20oW , and of structure B toward N350W(magnetic) .

Nt sw

2mo---.-.:':'.:'

.~. .-' . :~.':- ;'" \': ..::.:::;

------- ...---- + +------- ----......---;:. --- <,------ . ------ + + +++++++++ + ~""::"-'- "' ''

+ + + + + + + + + + ~ -......-......

+ + + + + + + + + + + + + + -----+ + + + + + + + + + + + + + --:::-----.......

+ + ++ + + .~

2. Diagram of southern face of body A at Mococa. Symbols: crosses, bedrock; open circles. conglome­rate; triangles, diamictite; unpatterned with discontinuous concentric lines, laminated fine sandstone;unpatterned, medium grained sandstone and soil; dotted.

NE.. 'wi'"~~--~----'"-----"'----"-------'''---''''---,..-''''--"'''------''''

3. Diagram of northern face of body A. Symbols as in fig. 2

NE sw

4. Diagram of southern face of body B. Symbols as in fig. 2.

sw

----- o

5. Diagram of northern face of body B. Symbols as In fig. 2. Darker oblique straight lineat lower right represent road surface.

50 BOL. DA SOCIEDADE BRASILEIRA DE GEOLOGIA - V. 17, NQ I, 1968

These structures are formed by a coreof the local basement rock which is granite­gneiss, covered by a veneer of diamictite,conglomerate and medium to coarse grainedsandstone. The structures are overlain bysiltstones and fine to medium grained felds­pathic sandstone (fig. 2-8). Locally ali thesedimentary rocks are red in color. Theentire sequence in roughly 6 meters in thick­ness. Another boss of granite-gneiss appearsin the southwestern extremity of northernface for which no counterpart was found onthe southern face.

Distribution of the sediments within eachbody varies between faces of the cut. Gene­rally the sediments tend to be thicker in thenorthern face. This is better shown forstructure A (figs. 2-3; 6-7) . The nor­thern face of body B is not well exposed,but presumably a thicker section of diamic­tite and/or conglomerate and sandstoneoccur below the road surface (figs. 4-5 and 8).The thickeness of the sediments varies alsoin relation to the core topography, being gre­ater in the lower parts and smaller in thehigher parts.

structure A has a single core (figs. 2-3;6-7) while in structure B, the granite-gneisscore is bifid, at least at the southern face(figs. 4 and 8); only the southwestern base­ment boss appears at the northern section ofstructure B (fig. 5).

Sediments forming the elongated bodiesshow a very high degree of compaction, es­pecially the diamictite that constitutes locallythe most resistent rock. The overlying rocksare friable and rest discordantly on theelongate bodies. Arching of the stratifica­tion over the structures indicates differentialcompaction of the overlying sediments (figs.2-5). Argillite laminae intercalated in sand­stone that cover the northern face of structureA are complexely folded in a manner sugges­ting slumping, diverging from the plane ofsymmetry of the body (fig. 3). Laminatesiltstones that appear in the upper part ofsouthern face of structure A seems to betruncated above the structure by the mediumto coarse sandstone bed (fig. 2) .

Both the surface of the granite-gneiss coreas well as the upper surface of the diamictiteof the two structure are normally smoothand frequently finely striated. The striae areparallel, with direction varying from N35°Wto N55°W (magnetic). (Fig. 9.)

The upper surface of the granite-gneisscore also shows irregularities formed byacute projections into the diamictite, underlainby horizontal wedge-shaped fissures filledwith diamictite. Vertical wedge-like cracksfilled with diamictite also occur (figs. 4 and10). At the northern face of structure B alarge block of granite-gneiss is a apparentlyonly slightly displaced from its former posi­tion separated from the basement core by athin film of diamictite (figs. 3 and 11).

Taking as a reference the plane of theroad bed, structure A has a section of 22meters at the southern face enlarging to 30meters at the northern face. The maximumexposed height also increases from 2,4 metersto 3,2 meters. Structure B shows a sectionof 32 meters at the southern face decreasingto 16 meters at the northern face. The heightalso diminishes from 2 to 1,3 meters.

Origin of the elongated topographic fonns

There are several indications that thestructures described formed bodies with reliefbefore the deposition of the overlying sedi­ments: the discordant contact, the attitude ofthe stratification of the overlying sedimentsindicating diferential compaction, the aparenttruncation of the beds towards the top, andthe slumping structures along the flanks ofthe bodies. Preservation of these featuresindicates rapid burial by the overlying rocks,before destruction by erosion.

Direct association with diamictite indica­tes that the features could have been formedby the streamline flow of the upper Paleozoicglaciers.

Features of glacial deposits of Pleistoceneage with characteristic of shape, structureand composition similar to the ones studiedat Mococa are assembled into a large groupcalled streamline molded forms by Flint(1957) . These include positive and nega­tive elements with dimensions varyingfrom centimeters (e . g. a glacial striae), tokilometers in length (e. g. the great drum­lins). Features of intermediate size, bothpositive and negative, can occur generally inlarge swarms, rarely isolated.

Among streamline molded forms the long­est and best known are the drumlins (Alden,1905; 1918; Slater, 1929; Hollingwort, 1931;~ravenor, 1953; Charlesworth, 1957; Flint,1957) . Drumlins, as well as other positiveelements, can have a composition ranging

A. C. ROCHA CAMPOS. J. E . S. FARJALLAT e R. YOSHIDA - New glacial . .. 51

6. View to the north east at Mococa showingsouthern face of structure A. Note thickerveneer of diamictite and sandstone.

7. View to the northwest at Mococa showingnorthern face of structure A. Lower whlteishzone marks contact of diamictite and basementrock and upper one marks contact of coveringrocks and diamictite.

8. View to the southeast at Mococa showing southern face of structure B. Note valley filled withconglomerate between the two granite bosses.

9. Parallel striae on the upper surface of dia­mictite at southern face of structure B (Mo­coca).

10. Ver tic al wedge-shaped cav ity in the bedrockfilled diamictite at southern face of structure B(Mococa) .

52 BDL. DA SOCIEDADE BRASILEIRA DE GEOLOGIA - V. 17, NQ I, 1968

11. Large granite-gneis block separated frombedrock by diamictite at northern face ofstructure A (Mo coca).

from entirely of bedrock at one extreme, toentire-ly of drift at the other. The origin ofthese structures is still a matter of debate.Bodies composed entirely of bedrock areobviously erosional. Those composed of driftassociated with a core of bedrock and, espe­cially those composed only of drift are moredifficult to explain. Essentially there aretwo ideas, either that they have formed byerosion; second that they have resulted fromaccretion of drift. Gravenor (1953) presen­ted a review of some theories concerning theorigin of drumlins and listed their mostimportan t characteristics.

A complete explanation for the origin ofthe features described in this paper presentsthe same problems.

At Mococa the elongated molded formcould have been formed by erosion, or erosionand deposition.

Polishing and striation of the upper sur­face of the granite-gneiss basement indicatesthat it was abraded by glacier ice .

The overlying diamictite is tightly packedagainst the irregularities of the basementtopography, filling the V-shaped cavities.This is interpreted as being caused by depo­sition under the weight of a glacier. Thepresence of striae on the upper surface ofdiamictite is indicative of erosion after, ormore probably during deposition of the veneerof sediments.

Whether the erosion of bedrock anddeposition of sediments was produced by the

same glacier or not is difficult to ascertain.However, the approximate agreement betweenthe elongation of the two bodies and the dire­ction of the striae on the basement and on thediamictite could be and indication that theerosion of basement and deposition anderosion of diamictite, conglomerate andsandstone occurred more or less simultan­eously.

The diamictite discordantly overlies theconglomerate filling the valley between thetwo granite bosses of structure B indicatingthat it was deposited by running melt waterbefore the accretion of the diamictite cover.

At structure A and possibly at structureB the thickening of the diamictite and sand­stone veneering the granite bosses suggest acrag-and-tail association.

The horizontal and wedge-shaped fissuresof the basement could be the effect of glacialplucking or they could represent frost cracks.Separation of the blocks of granite-gneiss inthe northern face of the structure could havehad a similar origins. Another possibleexplanation is to interpret these features asdue to sheeting of the granite-gneiss withenlargement of the fissures during depositionof the diamictite. This case would involveexposure of the granite-gneiss surface toweathering for a considerable space of time

before deposition of diamictite.In the Pleistocene deposits these elon­

gated molded bodies generally occur inswarms. The rarity of occurrences of thestructures in pre-Pleistocene sediments could,however, be explained by difficulties in pre­servation, especially of the larger bodies, andby lack of outcrops that make their recogni­tion difficult, or both.

Nevertheless, other possible elongatedmolded forms were examined in the Paranabasin. A good example is the one exposedin the two faces of a cut at kilometer 107 ofthe Sorocabana Railroad, near the town ofSorocaba.

, The structure has a ~hape similar to theone at Mococa, but ' it ill composed entirelyof sediments, mainly feldspathic, poorly sor­ted sandstones and mudstone (figs. 12-13).

At this locality there is no direct asso­ciation with diamictite. However, diamictitewith facetted and striated clasts and showingtillitic fabric is beautifully exposed some 200meters to the south west, probably only afew meters above the structure. The sedi-

A. C. ROCHA CAMPOS, J. E. S. FARJALLAT e R. YOSHIDA - New glacial.. . 53

12. View of southern face elongated structure at Sorocaba. Note truncation of overlying sandstone bedsover the structure and concentric discontinuous joints of the "core".

ments belong to the basal portion of theItarare Subgroup in the area.

An origin through folding for this featureis improbable since it is included in a flatlying sedimentary sequence. The apparentconcentric structure shown by the disconti­nuous concentric jointing of the «cor e» doesnot seem to be related to stratification.

Erosion and/or deposition by glacier icecould be responsible for the formation of theelongated body at Sorocaba. The concentricstructure parallel to its upper surface isremindful of the concentric banding describedby Slater (1929) in Pleistocene drumlins andconsidered indicative of origin through accre­tion of drift.

Clast concentration

The locality studied for this feature isthe two faces of a cut on the SorocabanaRailroad approximately 200 meters south ofthe railroad station at Jumirim. A briefdescription of thi~ outcrop appeared in Frakesand Crowell (1969).

In the roughly southeast-northwest cut,about 15 meters apart, a linear and horizontalconcentration of clasts occurs within the lower

part of a light olive-gray to dusky-yellow,medium-gray, indurated diamictite. The dia­mictite is underlain by me dium to corsegrained yellow-gray feldspathic sandstone.The upper part of the section is covered byvegetation (figs. 14-16) .

The clasts of the concent ra t ion vary insize from cobbles to blocks and boulders, upto 90 centimeters in diameter. The compo­sition of the clasts is the same as the rela­tively rare clast s dispersed in the diamictite,

13. View of northern face of elongated s t ruct u reat Sorocaba. Band of vegetation marks its upperlimit.

SE titil'-

NW

:<:.....;--1

~cut at ,!-"

.....<0

&5

~~~~"-""e>

15. Diag ram of so u th ern face of railroadJumirim. Symbols as in fig . 14.

~%. •

€ f :i.~:~..

SE

diP

2m

i~Iw.~mt~J,,:~

o

aodIO:-"",

tA1o!:'"'

~tno8t<JtJ>~tA1~

>tn......r­t<J......

~

~ot<J

Symbols : umpatterned, di amictite ; obliquely hatch ed , clast s ; stippl ed , sands tone ; squares , ca lcare ou s sand- 0line represent probable prolongation of u pp er intercala tion zon e; low er interprupted line represents boundary S

o......>

e(;G

. . ..- - - -~- - - - ... c-.. • •. -- - --- .. --.....

~ '

....GJ ~ OdI 0ll:l CP Gl) ~

.~~-

<99

- --- ~-- ~ ----~ ----- - -o -....;;;;;:; _ _ _

659 -

,Glb

e

----- -

. ! [~;~:fh:iWM{-;:;iW;~tl;·~iX(:~:![Y;~X~~~\~·1 ;' .

NW

_0

o

D ia gram of northern fa ce of railroad cu t at Jumirim.s t one; irregular dots, veget ation. Up per descont inu ousb etween the two zones of diamictite.

~.. . <, - . - - - - -~.~-..I!~~;..~~~-'====~~""";::;:=~~§5lb::===z::2>~~ e.. 'Ib c> '" ~ ...

(JP> .. D ... WlJP> q;, "" 1'1> _e 0"".;. .. ~. co' .. .... ~ '" (fJb., '"e~ ..~*· ·· ·· ,·:·~~~¥K~~~~:c~~-- - - ~- -,-- -~~~~':: ~ .:~- ~-~ _ .~--- ---.:::-- --~ - - - ~

~

1 --- __ --

14,

A. C. ROCHA CAMPOS, J. E. S. FARJALLAT e R. YOSHIDA - New glaciaL.. 55

16. View of the southern face at Jumirim. Note the two zones of the diamictite, clast concentration,and calcareous concretions at left.

17. Part of the clast concentration included indiamictite at the northern face at Jumirim.Note upper beveled surface of clasts.

which are predominantly quartzite and gra­nite-gneiss. Size variation and general shapealso coincide (fig. 14-15).

The horizontal concentration can befollowed for 50 meters along the betterexposed western face (fig. 14). At itssouthern part a higher dispersion of theclasts makes the recognition of the concen­tration difficult, but in the rest of the expo­sure the linear disposition is clearer, althoughin places the clasts are separated laterallyby 2-3 meters. At some places they arealmost in contact with each other with afew showing an imbricated disposition (figs.14-15). Direction of imbrication is approxi­mately toward N550W (magnetic).

At the southern face the concentrationis exposed over a distance of at least 12meters (figs. 15-16). The northwestern sideof the outcrop is covered by vegetation andtalus debris. The general characteristics arethe same as on the other face.

18. Layers and veins of sandstone showing celulardisposition within the diamictite on the nor­thern face at Jumirim.

Several clasts of the concentrationshow their upper surface facetted (fig. 17)bearing series of striae or narrow groovesoriented in a parallel fashion. Measurementof the striae in at least 10 clasts indicate onlyslight divergence, the average being appro­ximately N600W (magnetic). Other featuresare associated with the striae in some of theclasts, especially on the upper surface oflargest boulder at the northwestern extremityof eastern face. This includes crescenticmarks and possibly chatter marks. Orienta­tion of the axis of the crescentic marks isparallel to the striae on the clasts. Thesefeatures have been described in detailelsewhere (Rocha-Campos et al., 1969).

In the southern side of the northern faceand in the exposed part of the southern facetwo zones can be observed in the diamictite.These are separated by an undulating andtransitional boundary. In the lower one,about I meter thick, the diamictite is yellow-

56 BOL. DA SOCIEDADE BRASILEIRA DE GEOLOGIA - V. 17, NQ I, 1968

gray, fragmenting in small, centimetric slabs.Epigenetic calcareous concretions are commonin this zone and show a horizontal disposition.A test in the field indicated that the matrixis also more calcareous. In its upper part thediamictite is highly compacted and medium­gray in color (fig. 15). The clast con­centration is included mainly in the upperzone of the diamictite in the northern cut(fig. 14). In the southern cut most of theclasts are enclosed in the lower zone whoseupper limit cuts the clast concentration intwo places (figs. 15-16).

The northern face differs also from the

southern face by having two horizontal and

parallel zones of fine to medium grained

sandstone and calcareous sandstone interca­

lations in the diamictite, 1 to 1.5 meters

apart, and varying in width from a few

milimeters up to one meter (fig. 14). The

intercalations occur as horizontal or oblique

layers or veins of fine to medium grained

sandstone. The latter are extremely rami­

fied in places forming a kind of cellular or

honeycomb structure (figs. 14 e 18). These

pass laterally to irregular masses or lensesof highly calcareous sandstone. Contactwith the enclosing diamictites is sharp andregular in most places. No internal structurecould be observed in the sandstone.

The two zones are almost continuousalong the northern face, except for the sou­thern end of the upper one (fig. 14). The clastconcentration is included between the twointercalation zones along most of the nor­thern face (fig. 14).

Origin of the clast concentration

Boulder pavements are a type lag con­centrate that results from the removal of thefine matrix of tills by wind or running meltwater after deglaciation of a locality. Rene­wed glaciation abrades the concentrate for­ming a more or less continuous pavement ofclasts with beveled upper surfaces, thatcommonly bear striae oriented in a parallelfashion. The pavement is usually covered bytill of different composition than the tillbelow the clasts. Boulder pavements can alsoresult from the rejuvenation of a glacierwith renewed subglacial erosion after depo­sition of a till, without the interference ofa deglaciation period. (Flint, 1957; 1961). In

this case the overlying and underlying tillmay be lithologically very similar.

Several of the characteristics of the clastconcentration at Jumirim are comparable tothose of the Pleistocene boulder pavements.These include the horizontal and more are lesslinear disposition of the clasts and the upperbeveled surface of the clasts bearing parallelstriations and other marks (Flint, 1957;1961) . Similarity of composition of theclasts of the concentration and thoseincluded in the till is also in agreement. Theliterature does not refer to imbrication ofclasts in Pleistocene pavements but theiroccurrence is probable.

The main differences are the greaterlateral and vertical dispersion of the clasts,and lack of a diamictite of different compo­sition overlying them.

Lack of detailed description of boulderpavements in the literature makes compari­son of small features difficult. Lateral se­paration of clasts seems to occur in thePleistocene examples. Flint (1957, p. 120,fig. 7-11) ilustrates a boulder pavement fromCanada where the clasts are sometimes 2-3meters apart. Slight vertical dispersion alsocan be observed in the photograph.

The apparent lack of a diamictite ofdifferent composition above the clast concen­tration in the Brazilian example couldperhaps be explained by the absence of adeglaciation period following the depositionof the clasts. The composition of the matrixof the diamictite above and below the clastconcentration could be very similar and mix­ture of the two materials would make diffi­cult the recognition of a physical break.

An alternative, but less probable expla­nation would be to interpret the clast con­centration at Jumirim as a type of fabricresulting from shearing action inside alodgement till. The parallel disposition of thestriae and other glacial marks on theclasts, and the extension of the clast concen­tration seem to mitigate against this.

A definite explanation for the formationof the two zones of sandstone lenses, layersand veins is lacking. Geometric aspects anddisposition of these features coincide withthose of ice veins and layers developed infrozen ground for which they could representpseudomorphs (Taber, 1943; Schafer, 1949).

The zonation observed in the diamictitecould be due to recent weathering with

A. C. ROCHA CAMPOS, J . E . S. FARJALLAT e R. YOSHIDA - New glaciaL .. 57

leaching of t he ca rbonate from the upperpart of the diamicti te and deposition in thesandstone in t er ca la t ions and as concretionsin t he lower part of the diamictite.

CONCLUSIONS

El ongated t opographic form s compose dof bed rock wi t h a veneer of diamictite', sand­st one and conglomerat e, and entirely ofsed im ents occur in the Itarare Subgroup .These features show analogi es to str eamlineelonga ted bodies pr oduced by the flo w of icedu ring Pleisto cen e glaciation.

The lin ear clas t concen t r ati on en closedin a diamictite a t Jumirlm also show severalpoints in com mon with the desc r iptions ofPleistocene boulder pavements and could havea similar mode of origin.

Occurrence of these features directlyasso ciated w it h or included in a sequen ce

that contains diamictites is indicative of aglacial origin for these rocks.

Elongation of the structures and directionof striae on the upper surface of the bedrockand diamictite at Mococa are parallel andtrend N45 0W (magnetic) . Possible crag­and-tail distribution of the drift on thebedrock could indicate displacement of upperPale-ozoic glaciers from the southeasternquadrant. Elongation of the axis of thestructure at Sorocaba is toward N150W

(magneti c ) .Directions of striae and grooves and of

the axes of crescentic marks on the uppersurfaces of clasts of the boulder pavementat Jumirim are parallel and oriented towardN60oW. A few clasts of the pavement areimbricated, direction of imbrication beeingtoward N40 oW. This is interpreted as ind i­cative of ice flow toward northwest in thearea.

FLINT, R. F .. - 1957 - Glacia l a nd Pleistocen eGeology, J ohn Wiley & Sons, Inc., New York.

FRAKES, L . A. and CROWE L L . J . C. - 1969 ­L ate P al eo zoic g lacia tion : I , Sou th Am er ica,Geo!. Soc. Am er, Bull. , v , 80, n . 6, p p . 1007-1042.

CH AR L ESWORT H , J . K. - 1957 - The Quater­nary Era wit h special refe rence t o its g laciation,v. I , p p. 389-403, E d w ard Arnold Ltd . , London .

AMARAL, S . E . d o - 1965 - Nova oco r rencia d er och a m outo n nee em Saito, SP. B o\. Soc . B r as .Geo logia, v . 14, n. 1 e 2, pp. 71-82.

GRAVENOR, C. P. 1953 - The origin ofdrumlins, Am . Jour. of Science , v . 251. n . 9.pp. 674-681.

H OLLINGWORTH, S. E. - 1931 - The gl ac ia t ionof w estern Edenside a nd the d rumlins o f Eden­side and the Solwa y basin , Geo!. Soc. LondonQuart. J ou r . , v . fIT, PP . 281-359.

MARTIN, H . - 1961 - The hypothesis, of co n ti ­ne n t a l drift in the ligh t of recent a d va nces o fgeo logica l k nowledg e in Brazil and in S outhW est Africa , Geo!. Soc. o f South Af ri ca .An exure t o v . L XIV.

R OCHA-CAMPOS , A. C. - 1967a - The Tubara oGrou p in the B razilian portion o f P arana basin(w it h a co n t ri b utio n by J . E . S . F arjalla t onMato Grosso a n d Goi a s ) i n J. J . Biga r ella, R.D. Becker e I. D. P into (Bds.) Problems inB razilian Gondwana Geol ogy, p p , 27-102. Inst.Geo!. U niv . Federal d o P arana , Curi ti ba.

R OCHA- CAMPOS . A. C. - 1967b - Novas ocorren­cia s de f6ss eis m a ri nhos no Grupo Tubara o emSao P a ul o e Santa Cata r in a , B ol. Soc. B ras.Geo!. v, 15, nv 4, pp. 5-13.

R OCHA-CAMPOS , A. C,; FARJALLAT, J. E . S,a nd YOSHIDA, R. - 1969 - Crescen ti c markson a Late P aleozoic g lacia l pavement in Sou­theaster n Brazil , Bull . Geo \. Soc. Am er.. v. 80,n. 6, pp. 1123-1126.

SCHAFER, J . P . - 1949 - Som e periglac ia l f ea­tures in Ce ntra l Mo ntana , J . Ge ol og y, v . 57, n.2, p p. 154-174.

SLAT ER, G. - 1929 - T h e structure of the d rum­li n s expos ed on the sou t h shore of Lake On­t ario, New York State Mus. Bull. 281, pp. 3-19,

TABER, S . - 1943 - P erennially frozen g ro undin Alaska , Geo!. Soc. Am. , Bull . v. &4 , n . 10,p .p. 1433-1548.

d rumlins ofGeo!. Sur vey

TheU . S.

Geo logical ev ide nce ofN a irn (Ed.) Descriptive

140-155. Inte r sci ence ,

FLINT, R. F . - 1961 ­co ld c limate in A. E . M.P al a eocli m atol ogy , pp ,New York .

BIBLIOGRAPHY

ALDEN , W . C. - 1905so u theaste rn W isc on sin,Bull. 273, pp. 25-27.

ALDEN, W. C. - 1918 - The Quater n ary geologyof sou t heastern Wiscon s in, with a chapter onthe older r ock format ion s, U. S. G eo!. Survey ,Prof. Paper 106.

FRAKES, L. A. ; FIGUEIREDO FILHO. P, M. deand FULFARO, V. J . - 1968 - F ossil es ke rsand associa ted fea tu res from P a rana ba s in ,Bra zil , J our. Sed . P etr. , v. 38, n. I , PP . 5-12,

BIGARELLA, J . J . ; SALAMUN I , R . and FUCK. R .A. - 1967 - Striated surfaces and r el a t edfeatures d eveloped by the Gondwana ice-sheets(Stat e of P arana, B razil), Pal,a eogeogra phy ,P al a eoclimatolog y , P al aeocology 13, art . 56. pp.265-276 .