Reprinted from THE Jounwer or Complnerrvn NnunolocyVol. 123, No. 1, August 1964

The Effects of crn Enriched Environment on theHistology of the Rcrt Cerebral Cortex

MARIAN C. DIAMOND, DAVID KRECH ,qno MARK R. ROSENZWEIG 1

Departments of Anatomg and Psychologg, Unitersity of California,Berkeley

Recently, it has been reported by Rosen-zweig, Krech, Bennett, and Diamond ('62)that increasing the environmental com-plexity and training of rats resulted inmeasureable changes in brain chemistryand in brain weight. More specificially,some of the results have demonstrated thatthe cerebral cortex from rats subjected totraining showed not only an increase inacetylcholinesterase but also an increase inweight. It is this weight increase whichconcerns us at present. Though small,7.6% in the samples from the visual cor-tex and 3.37o in the samples from thesomatosensory cortex, the increase wasconsistent and repeatable since the corticalweights from trained animals exceededthose from untrained and isolated rats in79% of the cases for the visual area (53of 67 pairs) and in 64% for the somato-sensory arca (43 of 67 pairs ) (Rosenzweig,Krech and Bennett, unpublished data). Incontrast to the cortical weight increase,there were no changes in weight in thesubcortex. Also it might be pointed outthat the body weights of the isolated ratswere greater than those of the trained ratsby approximately 7Vo .

Heretofore the brain has generally beenconsidered incapable of rnacroscopic phys-ical changes as a consequence of altera-tions in experience. Because of the find-ings cited above, anatomical examinationswere considered essential to determine pos-sible causes for this increase in corticalweight.

METHODS

Behaaioral conditionsIn Experiment I, 11 pairs of male litter-

mate rats from the Sr, maze-bright strain(U. C.Psychology Colony) were used. Theentire experiment was replicated (Experi-ment II) with nine pairs of animals of the

J. Cour. Nnun., 123; 111-120.

Sr strain. The behavioral procedure hasbeen described in detail by Krech et al.(Krech, Rosenzweig and Bennett, '60), butwill be mentioned briefly here.

At 25 days post partum one animal ofeach pair, chosen at random, was placed inthe ECT (Environmental Complexity andTraining) group, the littermate being as-signed to the IC (Isolated Condition) group.The ECT animals were housed tosether ina large cage (25" by 25" by 18t) . Twodifferent wooden "toys" from a set of sevenwere put in the cage each day. There wasalso a small wooden rnaze in the casewhich the rats used as a nesting box. For30 minutes each day the rats were allowedto explore the Hebb-Williams rnaze,with the pattern of barriers changed daily.At about 50 days of age formal maze train-ing began in the Lashley III maze, theDashiell rnaze, and the Krech HypothesisApparatus. Glucose pellet rewards weregiven. Food and water were normallyavailable ad libitum in the home case.

At 25 days of age the l ittermate of theECT rat was placed in the IC group, con-sisting of an individual cage (1,I" by 8" by8") with solid metal walls on three sides.Food and water were available ad libitum,and a glu-cose pellet was given each time anECT animal was so rewarded. The animalsdid not see or touch another animal.Neither were they handled during clean-ing of the cages. All of the animals in bothgroups were weighed at weekly intervals.Both the ECT and IC groups remained

1 This investigation was supported in part by Re-search Crant G-20895 to D. Krech, M. R. Rosenzweis,and E. L. Bennett , and Research Grant G-2542 toD. Krech and Marian C. Diamond from the NationalScience Foundation; in part by Research Grant M-1292to D, Krech, M. R. Rosenzweig, and E. L. Bennettfrom the National Institute of Mental Health, UnitedStates Public Health Service; and in part by ResearchContract DA-49-193-MD-2329 to D. Krech, M. R. Rosen-zweig, and E, L. Bennett from Surgeon General's Office.It also received support from the United States AtomicEnergy Commission.

111

772

about 80 days in their respective experi-mental conditions.

Anatomical methodsAt the end of the experimental period,

the animals, now about 105 days of age,were brought to the histology laboratory forsacrifice. The previous treatment of eachindividual animal was not made known tothe histologists; only animal numbers weregiven. The animals -,vere decapitated, thebrains removed and placed in 1O% formol-saline. Transverse frozen sections were cutat 25 u. To be certain that cortical sec-tions were taken from a similar region ofthe brain f rom each animal, subcort icallandmarks were used to identify the spe-cific areas. In removing a sample from thesomatosensory cortex, the anterior com-missure served as a guide. Where ther ight and lef t anter ior cbmmissures decus-sated in the midline, four sections weretaken. Two were stained with Windle'smodified Nissl stain (Windle. Rhines and

MARIAN C. DIAMOND, DAVID KRECH AND MARK R. ROSENZWEIG

Rankin, '43), and the others with hema-toxylin and eosin.

The appearance of the cisterna magnaseparated from the aqueduct of Sylvius bythe posterior commissure was the land-mark at which four visual cortical sectionswere taken. These tissues were stained ina manner identical with the somatosensorysamples.

Sections stained by Windle's bufferedthionine solution were used for measuringthe cortical depths and for differential cellcounts. In calculating the depth of thecortex, eight readings with an optical mi-crometer were taken on one her:nispherefrom the dorsal aspect of Layer XI to theunderlying white matter. The first readingwas consistently taken immediately lateralto the elevation of the corpus caliosum.The succeeding readings were madeone micrometer width (150 u) lateralto each prerrious measurement until eightconsecutive measures were completed (fig.1 ). Also on additional sections cortical

4Itrl

i:*.,

I

-r..:r

iifffi#,'$'*ti,::l:

i ,.rt1'"i;,vi.,1i\ ;

{1

\qi-,git+r-rt,. .:. : ", ,: _:i l ' ' l 11i':'x.r.}ilj|:

. , , ;

Transverse section throughmicrometer readings through

- 6- , - { .

Fig. 1tions of

the rat somatosensory cortex. Numbered lines indicate posi-the cortical layers to measule the cortical depth,

depths were measured on both right andleft hemispheres. On these latter sections,Layer I was included in the depth meas-urements and was also measured sepa-rately.

Differential cell counts were read with anoil immersion lens on the same slideswhich were used for measuring the cor-tical depths on one hemispherel The cellcounts were taken at the same position aswas the first micrometer reading, i.e. im-mediately lateral to the elevation of thecorpus callosum. The counts included onlythose neurons with nucleoli. Differentialglial counts were made from the nuclei ofoligodendrocytes, astrocytes, and microglia.

Blood vessel counts were made underhigh dry magnification (400 X). All sizesof blood vessels were clearly defined withhematoxylin and eosin, for even if, attimes, no endothelial cells were present,red blood cells in a continuous line indi-cated the presence of a vessel. An Indiaink-gelatin mixture was used in a prelimi-nary attempt to demonstrate blood vessels.This method was abandoned, for it was feltthat under the pressure of the injection,some vessels were being expanded beyonda normal picture. Also the pressure of theinjections could not be standardized to thepoint where all animals were receivingidentical treatment.

For counting, the vessels were dividedinto two groups, a diameter of 5 u being thedemarcation line. This division was chosenin agreement with the work of Eayrs ('54)who arbitrarily selected 5 u to be the upperlimit of size for capillaries in the rat. Thefirst reading was taken from the surface ofthe cortex throush Layer VI at the positionimmediately late"ral to the elevation of thecorpus callosum. Three additional read-ings were made one microscopic field lat-eral to each previous reading.

113

RESULTS

Cortical depthThe mean depth of the visual cortex was

greater by 6.27o in the ECT than in theIC animals in both experiments I and II(table 1). Within each experiment, sig-nificance was calculated by t-test for cor-related data. An analysis of variancerevealed that the results of the two experi-ments did not differ from each other sig-nificantly and that the overall ECT-IC di}-ference of. 6,2Vo was highly significant(P > 0.001). The ECT effect was some-what greater in the right than in the lefthemisphere in both experiments (table 2),but this interhemispheric difference wasnot significant.

Layer I was included in some measure-ments (table 2) and not in others (table 1),for the pial surface on some sections wasat times distorted, making measurementsinaccurate on these tissues, However, withthe pial surface intact, the depth of LayerI was measured with high accuracy. It wasfound that for Layer I the IC mean actuallyexceeded the ECT mean, but only by l.7Vo ,and the IC animal exceeded its ECT litter-mate on this measure in 11 cases out of20. We may, therefore, conclude thatLayer I was not concerned in the depth in-crease in the visual area.

In the measurements of weight differ-ences, the somatosensory cortex had notshown as large or clear cut results as thevisual region, and this was found again inthe present measurements of corticaldepth. There was greater variability insomatosensory depths, both among ani-mals in a single experiment and betweenexperiments, than in the visual depths.In the first experiment the ECT depthswere greater than the IC depths; in thesecond experiment, the reverse was true.

ENVIRONMENTAL EFFECTS ON RAT CORTEX

TABLE 1

Depths (p) of ztisual cortex frorn ECT ond IC rats (excluiling Lager I)

Experiment Mean 5. E. Mean s.E. ECT/IC ECT vs. IC ECT > IC

I

II

I+I I

119

20

133214041364

L72916

7277 221298 277284 t7

1.0481.0821.O62

< 0.001 7 / t r< 0.001 9/9< 0.001 76/20

1.r4 MARIAN C. DIAMOND, DAVID KRECH AND MARK R. ROSENZWEIG

TABLE 2

Depths (p) of oisual cortex frotn ECT and IC rats indicating differences betz4)eenright and. Ieft hemispheres (including lager I)

Experi-ment

Hemi-sphere

ECTMean

ICMeanN

ECT/IC ECT vs. IC ECT > IC

I

I

I

9

I I

II

1 l

11

11

R

L

R+L--2-

I

R+L

- 2

R+r,-'-z

1323

L32L

7322

1300

1304

1.302

13 13

L270

1280

L275

1238

1250

1.244

7260

8/7L

9/7L

9/LL

7/9

7/9

7/9

76/20

7.042 < 0.05

1.032 < 0.01

1.037 < 0.01

7.049 < 0.02

L.O42 < O.05

L.O46 < 0.02

L.O47 < 0.001

II

I+I I 20

TABLE 3

Depths (p) of somatosensorA cortex frotn ECT and. IC rats (excluding kryer I)

ECT ICExperi-ment N

Mean S.E. ECT/IC ECT vs. IC ECT > IC

IT

I+I I

1.96219751968

t6zD

19851896

7.O750.9951.038

22382t

184020

108

l8

< 0.001 7o/7ONS 3/8

< 0.01 73/78

TABLE 4

Mean nutnbers of neurons rter rniscroscopic field in oisual cortex in ECT and IC groups

ECTExperi-ment Mean S.E. Mean S.E. ECT/IC ECT vs. IC ECT > IC

III

r+i l

3.683.183.45

10I

18

o.22o.26o.77

0.8260.8390.831

4.45 0.253.79 0.244.15 0.18

< 0.05 2/7ONS 7/8

< 0.o1 3/18

Combining experiments I and II showedthe ECT mean greater than the IC meanby 38% (P < 0.01) ( table 3). Herethere was no consistent difference betweenthe depths of cortex of the right and lefthemispheres. The depth of Layer I wasmeasured in the somatosensory area, butthe inconsistent results prevent a definitereport at this time.

CeII countsThe average number of cells was counted

in each microscopic field, reading verticallyfrom the pial surface to the underlyingwhite matter.

In the visual area (table 4), the neuronsper microscopic field were less numerousin ECT than in IC animals in both experi-ments. The overall difference for both ex_periments amounted to ITVo and wasstatistically significant (p < 0.01). In thesomatosensory area, the ECT group had7Va less neurons per microscopic netd torexperiments I and II combined (p < 0.05)(table 5). The decreased numbers of neu_totas per fleld in the ECT animals were to beexpected in terms of their increased depthof cortex, but the magnitudes of the dif-ferences in neuronal dlnsity were severaltimes as large as the differences in cortical

ENVIRONMENTAL EFFECTS ON RAT CORTEX

TABLE 5

Mean numbers of neurons per microscopic field in so'matosensor?l cortex in ECT and IC groups

Experi-ment Mean ECT/IC ECT vs. IC ECT > IC

115

s.E.s.E.

III

I+I I

3.7L3.203.49

0.280.180.18

0.9340.9140.926

118

19

3.973.50o. l I

0.30o.200.20

NS 4.5/ tLNS s.O/8

< 0.05 7.5/19

depth. Again the differences were largerin the visual than in the somatosensoryarea.

In examining the distribution of neu-rons with nucleoli in the visual area, the ICanimals showed more cells per field thanECT animals in Layers II, III, IVb, c, andVa. In Layers IVa and Vb the ECT countswere equal or more numerous than the ICcounts ( f igs.2 and 3).

In the somatosensory region the IC ani-mals had more neurons per field than theECTs in Layers II, ry and V. Layers IIIand VI were less consistent, with ECTsand ICs reversing positions throughoutthese layers (figs. 4 and 5).

The mean number of glia per micro-scopic field in the visual area was less by6.77o in the ECT than in the IC animalsfor experiments I and II combined (meansof 2.8 and 3.0 glia per field for the twogroups, respectively). In the somatosen-sory area they were less by 5.9% in theECT (3.2 glia per field) than in the ICgroup (3.4 glia per field). None of theglial differences was significant.

In the cortical layers, the difference indistribution of glia was so small that therewas a continual reversal between ECT andIC groups, but with the overall mean forthe ECT group being slightly less. It is tobe noted that the somatosensory area had

200 400 buu aalr)

DEPTH (P)

Fig. 2 Mean distribution of neurons in rat visual cortex, Experiment I.

Jultra^z.or) ,LrJ az.

6z

-iz

oJhJ

f,(f)

zolUzL

z,

4

- ; /_,! / .

Y'

,--^n

,^-'.-----a \

ItrE

N =B

?--r , ra

tt-

E

7#v

olCr ECT

-;>d8d

E

200 4@ 800 tooo 1200 r4oDEPTH (z)

Fig. 3 Mean distribution of neurons in rat visual cortex, Experiment

'o"

116 MARIAN C, DIAMOND, DAVID KRECH AND MARK R. ROSENZWEIG

oJUtr@zoEl

zL

oz

oJU

azlrluzu

d

z

2d,'/I

rI

]I

\ - .H

m E

N.l l

i,^\ -o--o\ " - .

\ o-r-

v

o lc. ECT

>"*4_31-+\u

Fig.

DEPTH (/)

4 Mean distribution of neurons in rat somatosensory cortex. Experiment I.

Fig. 5 Mean distribution of neurons in

more glia per field than did the visual areaby 14%; the ECT and IC groups did notdiffer in this respect.

In the Sr strain of rats the ratio of gliato neurons in the visual and somatosensoryareas of the cortex was approximately oneto one. The ECT group has a smaller neu-ron/glial ratio than does the IC group inboth the visual and somatosensory samplesin experiments I and II (tables 6 and 7).

rat somatosensory cortex, Experiment II.

Since the total number of neurons is pre-sumed to be fixed, this indicates an in-crease in the total number of glia in theECT group.

Capillary countsThe number of capillaries per field was

less in the ECT group than in the IC groupby 72% (P < 0.05) in the visual area andby 8% in the somatosensory area. The

TABLE 6

Neuron/glial r&tios per rnicroscoTtic field. in aisual cortex in ECT and IC groups

il, I

d/J

tr Itr

-4

E

N =8

t.^

V/r.{v

olC. EcT

--"{rlAu

80DEPTH (p)

ECT ICExperi-

ment S.E. Mean s.E. ECT/IC ECTvs. IC ECT > IC

III

I+I I

L.28l -zo7.27

1.531.331.44

0.180"06o.10

0.8374.9470.882

0.09o.11o.o7

108

18

NS 2/10NS 2/8NS 4/r8

TABLE 7

Neuron/glial ratios per tnicroscoTtic fteld in soTnatosensorA cortex in ECT anil lC groups

ECT ICExperi-runt N

Mean S,E. Mean S.E. ECT/IC ECT vs. IC ECT > IC

II]

r+i l

1.18

19

1..18 0.100.97 0.o41.O9 0.06

3..23 0.101.01 0.o41.11 0.06

0.959o.9600.956

5.5/1.L3/8

8.s/ t9

NSNSNS

ENVIRONMENTAL EFFECTS ON RAT CORTEX

TABLE 8

Blood tessels Tter f i.eld in the aisual cortex from ECT and IC rats in experi 'ment I

t17

Type ofvessel Mean S.E. ECT/IC ECT vs. IC ECT > ICMean

CapillaryOver 5 lr

6.772.39

0.28U.JI) r..90

0.32o.22

0.873 < 0.O51.258 < 0.05

1/97/s

TABLE 9

Blood z;essels per field in the somatosensory cortex fro'rn ECT and IC rats in exTterirnent I

Type ofvessel S.E. s.E. ECT/IC ECT vs. IC ECT > IC

CapillaryOver 5 p

8.002.89

0.540.36

NS 2/7NS 6/8

d.b /

2.83o.720.29

0.9231.021

number of vessels over 5 u in diameter inthe ECT group was increased greatly overthe IC group in the visual area - 267o(P < 0.05) - and oniy slightly in the so-matosensory area - 2% . The overall vas-cularity of the somatosensory area wasgreater than in the visual area (tables 8and 9).

DISCUSSION

These results indicate that a complexenvironment leads to measureable histo-logical changes in the rat cerebral cortex.The depth of the cortex increases withenriched experience, and the number ofcells per field decreases. These changesare more marked in the visual area thanin the somatosensory area. Since the depthincrease in the visual area amounts to6% and the neuron density decrease to17 7o . it is not unreasonable to assumethat the cortex of the visual region isexpanding in volume and not in one direc-tion only. (An increase of 60/o in lineardimension would be expected to yield anincrease of 797o in volume.) In the morevariable somatosensory area the correlationis not as good. The depth increases by4Vo , and the density of neurons decreasesby 7%.

The expansion of the cortex might bedue in part to further development ofdendritic branching, a possibility supportedby the following reports of other investi-gators. Allen ('12) and Sugita ('17) re-ported that the rat cerebral cortex has afull complement of neurons 20 days after

birth. Any subsequent development is byperipheral extension of dendrites and in-creased amount of branching. Eayrs andGoodhead ('59) showed that in the rat thecortex has attained adult characteristics by18 days, and the average number of den-drites arising from the perikaryon hasreached the adult figure as early as thetwelfth day of age, further developmentbeing due to dendritic branching. In favorof dendritic ramifications being respon-sible for increases in brain weight are thereports of Erp Taalman Kip ('38), Horn('55), and Tower (Tower and Ell iott, '52)who found a highly significant correlationbetween brain weight and fiber count, indi-cating that brain weight is directly relatedto fiber density. However, Layer I, beingrich in dendritic branches, is not affectedin the visual area by the enriched enyiron-ment. According to Eayrs and Goodhead('59), Layer I shows very l itt le changeduring the growth of the rat cerebralcortex.

Studies are under way at present to de-termine by the Golgi-Cox method whetherincreased dendritic ramifications in theECT group are the major components re-sponsible for the depth increase. Thiswork is being carried out by Dr. P. CoIe-man at the University of Maryland.

The greatest difference between cellcounts in the ECT and IC groups is ap-parent in Layers II and III. According toIfuieg ('46), the main mass of intercellularmaterial in the supragranular layers isdendritic, again giving support to the hy-

118

pothesis that the ECT cortical increase isdue in part to dendritic branching.

The indication of an increase in totalnumber of glia with ECT is consistent withrecent measures of acetylcholinesterase(AChE) and cholinesterase (ChE) in thecortex of the ECT and IC rats (Bennett,Krech and Rosenzweig,'63). AChE, whichoccurs chiefly in neurons, decreases perunit of tissue weight with ECT, whileChE, which occurs chiefly in glia, increasesper unit of weight with ECT.

The present results indicate that thesomatosensory area has a greater vascu-Iarity than does the visual area. a findingin accord wi th that of Craigie ( '21). Thaithe ECT group shows less capillaries perfield, is in agreement with our findings ofa smaller number of neurons tler micro-scopic field. indicating again tl iat the in-crease in the depth of the cortex is duelikely to an increase in the dendriticbranching between the cells and capil-laries. The ECT group shows an increasein larger vessels, indicating a demand fora greater blood supply. Since the largervessels (those over 5 u) are running radi-ally for the most part, their increase innumber would not affect the cortical depthas measured. In order that measures ofcell density would not be affected by thepresence of blood vessels, the area to beincluded for cell counts was first examinedunder low power magnification to makecertain no large vessel-s would be encoun-tered durins the final cell counts under oilof immersion.

That there are more larger vessels inthe ECT group and less smaller ones mightbe explained in the following way. Per-haps all of the blood vessels in the ECTgroup increase in size, but due to the thin-ness of the small capillary walls thesevessels collapse after death and leave noindication of previous size. This statementis in asreement with de Vries ('34) who re-ported that brain tissue exerts pressure oncapillaries to compress them after death.On the other hand, the larger vessels witha thicker wall remain distended and moreof them can be accounted for in the ex-perimental group. Thus, the ECT groupshows more larger vessels and fewersmaller vessels than their IC littermates.Without considerins the possibilitv of ves-

MARIAN C. DIAMOND, DAVID KRECH AND MARK R. ROSENZWEIG

sels collapsing, these results could be ex-plained by the hypothesis that all vesselsincrease in size with the ECT condition.Thus, more vessels are counted in the largevessel (over 5 u) category, and less in thesmall vessel catego-ry.

It is well known that the degree ofvascularity corresponds directly w-ith thedegree of functional activity, Craigie ('38),Mott ( '14), and Scharrer ( '45). Further-more, with specific reference to the nervoussystem, Dunning and Wolff ( '37) havereported that the vascularity of tissuesvaried with the number of synaptic struc-tures and not with quantitative differencesin nerve cell bodies. Since we have arguedthat the increase in cortical depth may bedue to additional dendritic ramifications,then an increased blood supply is not un-likely in the ECT brains.

It is necessary to mention that increaseddendritic rarnifications are, undoubtedly,not the only factors to be considered aspossible causes for the cortical depth in-crease. The possibility of changes in suchfactors as perikaryon size, glial volume,cortical water, sodium, potassium, andchloride content, should not be dis-regarded.

SUMMARY

1. An enriched environment increasesthe depth of the cortex in the rat brain.

2. The depth increase is greater in thevisual cortex than in the somatosensorycortex.

3. The effect of the enriched environ-ment is evident in both the risht andleft hemispheres in the visual area. Inthe somatosensory area the differences arenot as consistent.

4. Layer I does not appear to be involvedin the depth increase in the visual cortex.Layers II and III show a greater increasein the ECT group than do ihe other layersin the visual area.

5. The neurons, glia, and capillaries areless numerous per fi.eld in the animalsexperiencing the enriched condition, indi-cating a greater amount of intercellularand intervascular substance. It is sug-gested that increased dendritic branchingiray partially account for this substance-.

6. There are more large vessels in theECT animals and less smaller ones; a pos-sible explanation for this result is offered.

ACKNOWLEDGMENTS

We are extremely grateful to Dr. GraceJackson and Mrs. Fay Law for their skilledand patient technical assistance. Wewould like to extend appreciation to Mrs.Carol Saslow for her help with the statis-tical analysis. We are indebted to MissMary Bennett for her histological analysis.

LITERATURE CITEDAllen, E. 7912 The cessation of mitosis in the

central nervous system of the albino rat. J.Comp. Neur., 22: 547-568.

Bennett, E. L., D. Krech and M. Rosenzweig1963 Effect of environmental complexity andtraining on acetylcholinesterase and cholines-terase activity in rat brain. Fed. Proc., 22: 334.

Craigie, E. H. 1938 Comparative anatomy andembryology of capillary bed of central nervoussystem. Res. Publ. Ass'n Nerv. Ment. Dis.,18: 3-28.

1921 Vascularity of the cerebral cortexof the albino rat. J. Comp. Neur., 33: 193-212.

De Vr ies, E. 1934 Psychiat . Neurol . , Bl . , Amst. ,Jaarg (Feestbundel Kappers), 712-722.

Dunning, H. S., and H. G. Wolff 1937 Therelative vascularity of various parts of thecentral and peripheral nervous system of thecat and its relation to function. J. Comp. Neur.,67: 43345O.

Ears, J. T. 1954 The vascularity of the cere-bral cortex in normal and cretinous rats. J.Anat. ,88; 164-173.

119

Eayrs, J. T., and B. Goodhead 1959 Postnataldevelopment of the cerebral cortex in the rat.J. Anat. , 93: 385-402.

Erp Taalman Kip, M. I. Van 1938 Licha-amsgrotte en hersenschors. P. Noordhoff,Groningen-B atavia.

Horn, G. 1955 Thyroid deficiencies and inani-tion. Anat. Rec., 121 : 63-79.

Krech, D., M. R. Rosenzweig and E. L. Bennett1.960 Effects of environmental complexity andtraining on brain chemistry. J. Comp. Physiol.Psychol . , 53; 509-519.

Krieg, W. J. S. 1946 Connections of the cerebralcortex. J. Comp. Neur., 84-85: 277-323.

Mott, F. W. 1914 Nature and nurture in mentaldevelopment. J. Murray, London, 151,

Rosenzweig, M. R., D. Krech and E. L. BennettUnpublished data.

Rosenzweig, M. R., D. Krech, E. L' Bennett andM. C. Diamond 1962 Effects of environmentalcomplexity and training on brain chemistryand anatomy: a replication and extension. J.Comp. Physiol. Psychol., 55: 429'437.

Scharrer, E. 1945 Capillaries and mitochondriain neuropil. J. Comp. Neur., 83; 237-243'

Sugita, N. 1917 Comparative studies on growthof cerebral cortex. J. Comp. Neur., 28: 495-591.

Tower. D. B., and K. A. C. Elliott 1952 Activitvof acetylcholine system in cerebral cortex ofvarious unanesthetized mammals. Amer. J.Physiol . , 168 : 747-759.

Windle, W., R. Rhines and J. Rankin 1943 ANissl method using bufiered solutions of thi-onin. Stain Technology, 78: 77-86.

ENVIRONMENTAL EFFECTS ON RAT CORTEX