"eves Diarrotd

right confir,crr llrnlted asnance rnigllror the well-

LXI ' r :RrNrEN.[AL NEr.JRolo(]y gg, l9g_101 ( I9g5)

On the Brain of a Scientist: Albert Einstein

MaRreN C. DIntroNn,* ARNoLD B. ScHerar l , fGneeR M. MuRpHy, Jn. , t aNn THouas HnRveyr

'Departments of Physiology'-Anatomy and {,4nthropology, tJnitersitl, of Catdornia, Berkelev,California 94720, and lDepartnents of Anatomy and psvchiarry, Ltniversitl, of

Caldornia, Los Angeles, Caldornia 90024

Received Seprember 25, 1984

Neuron:glial ratios were determined in specific regions of Albert Einstein's cerebralcortex to compare with samples from I I human male cortices. cell counrs weremade on either 6- or 20-pm sections from areas 9 and 39 from each hemisphere.All sections were stained with the Kliiver-Barrera stain to differentiate neurons fromglia' both astrocytes and oliogdendrocytes. Cell coun6 were made under oil immenionfrom the crown of the gyrus to the white matter by foilowing a red line drawn onthe coverslip. The average number of neurons and glial cells was delermined permicroscopic field. The results ofthe analysis suggest that in left area J9. the neuronal:g l ia l rat io for the Einstein brain is s igni f icant ly smal ler than the mean for the conrrolpopulat ion (r = 2.62, df 9, p < 0.05, two-tai led). Einstein 's brain did not di f fers ignihcanr ly in the neuronal :$ ia l rat io f rom the controls in any of the other rhreeareas studied. c) t98i Academrc pre$. Inc.

INTRODUCTION

Albert Einstein is generally conceded to have had one of the greatestscientif ic minds that ever existed. whereas neuroscientists may have noidea what characterized the brains of an Aristotle, Gali leo, or Newton asidefrom the extraordinary quality and prodigous quantity of their work, weare fortunate when we turn to a consideration of Einstein. we recently hadthe privilege of access to sufficient tissue from Einstein's brain to makecertain quantitative measures. Because of the method used for preparingthe tissue for histological examination, we were l imited in the kind ofanalysis we could make.

rAppreciat ion is extended to Ruth E. Johnson and E. Rosal ie Greer, ph.D, lor therr excel lcnttechnical assislance and to Doug Coe for his editorial critique. Dr. Harvey's current address isWcston, MO 64098.

0014-4886/85 $3.00Cop!n8ht !c, 1985 by Acadcmic Press. Inc.

{ l l nghLs dl rrproduclron rn anv form re*ned

198

EINSTEIN'S BRAIN CELLS

METHODS .

199The French mathematic ian, Jacques Hadamard, was interested in deter_mining rhe narure of rhe menrai ; ; ; ; f marhemaric ians. He conducreda psychorogicar survey of the

-mentar--images or internar words whichmathematic ians use , . .whether

they

# n:"; ;y.: i i, I ",ii' o b' H; ;; ;;'J'' l' ili'' X'li?l'. l' li il ";.H:" : ffH e re r t, n ", :. "'. ","1,i:: h" ", #;'"? ?J:[' l, JT IT T:'.'..":i :f : ti#tT:; ;::H."J:r:.,r.

.rr"n tiur i*i" r., .i "

prod u cri ve thou gh t. Ei nsrei nwas rhe emotionar 0""1"":]:t

t: i,-ve

finallv "at logicaily .onn..lJ conceprstypes', (10).

sls ol a rather vague play of visuai and ,o-. -rr.utu.It is doubtful whether any singre region of the brain mediates a, crossmodal interactions- When studyi;;;i;;Jr,s brain, *" i;;;;;..,...rru.,to choose regions *rr:h seemed";;-'"f;;*.the read provided by hisrntrospection' we decided-to examin" .orti.ur association regions of thesuperior prefrontal and inferior o".i.i"r'iri.s in the rigrt and reft hemi_spheres. Neuronal:gtial rarios i" ,1.r. *.";;:.:"ii'":,,:fl ::: ],one varid

-.uru..-oilhe srarus

"rr*r"r"r ffi:ff'ilj:ted as representing

' ' l

A contror base of mare human brains had, been obtained during the lastfew years from the Vereran's ea-i"lt"lii* Hospirar in Martinez, California.These incruded r r brains non,' inoiuiJurr, o, ,o g0 years of age who haddied from nonneurorogicattv .etatJ otr."ril tn" "*."*.;;:, 64 years;

"r'ff:;ll ;:: J:":,, :, " trrn " " r J."irll'rc,i,o n or ogica r "r1 r,-, " i I ecessa n I ya r s o p r a y ",i ;; ;; ;if ,._11 :ltr, ff l"t?'":'J:,X : ru: ::rm,_ ..;"":

ff#:fi:.t5}'J"*fl 'J; *i'', i' u- in''J"i -'

n' i s that,n., lo n ot co meFrom the Formarin-fixed brains-of former vA Hospital patients, brocksof cerebrar cortex about r .25 cm2 *.r.

-r.-oued from area g (superiorfronrar gyrus on

'n"..^oi:u'.ru,.*i ,ur"ro..) uno area 3g (inferior parietai

ii* il,:x' : r ?,' :' il: :T l #;;j, it, : "' ", ". ev -,, "." "-" d i n g, h ehemispheres G.;;t: r). The brocks *...,tut'

from both right and leftrhe surrace u' po,,ifri. ll9 .d*prv "no*i ;Ji;:,;l"j:# fff;ild;.J3;,I:Iralrc:.

Of .rhis group of I I brains. n.r.i ,.rrom 8 brai ns a"nd .lrroioin *.,i""r'"il ;'H: ffi ::: [TJ:T'::::. The Einstein brain blocks f.orn u..u, 6'lh em isph eres

".i ".J' i " rhe raboraroo ;,.;" j;0.;1J:::,,1. .:fi ::.il: J;:micrometer sections were taken non' einst.rn,s brain. Four to six secrions

, jl r ll t ll r i

J"I

200 DIAIVIOND ET AL,

Frc. l. A lateral view of the human brain indicating the position of the samples removed

for cell counls. A represents the sample from area 9 and B, area 39.

were cut from each block, Einstein's and the controls'. All brain sections

were stained with the Kliiver-Barrera, luxol fast blue cresyl echt violet

stain, to differentiate neurons from glia. After staining, one of the six

sections from each block was chosen for study. To assure the vertical

orientation of the cell counts, a straight l ine, perpendicular to the crown of

the gyrus, was drawn with a red pen on the coverslip. This ruled l ine was

kept just out of the field of vision as the cell counts were made, beginning

at layer II and extending into the subcortical white matter. Cell counts were

made with the aid of an oi l immersion lens (100X) and an eyepiece ( l0X),

with a ruled graticule placed in the eyepiece.

Since the demarkation between cortical gray matter and the underlying

white is not as clear in the human brain as in the rodent brain (3, 5), the

number of microscopic fields sampled was more arbitrary. Counts were

made into the white-gray boundary for one or two fields depending on the

density of the myelinated fibers, which are clearly demonstrated with this

stain. Two vertical columns were counted in the brain sections and the

average number of both neurons and glial cells per microscopic held was

determined.

I.4I; i

liffiiffit

ifrHflrl

$'flfrflflit,liltflxli!

1lfltr

il, {

I

EINSTEIN'S BR.\ IN CELLS 201

l

The counts were made in the fo l rowing manner: Beginning at the junct ionof layer I with layer II, the purple-stained neurons with crearry definednuclei and nucleoli wer_e counted in a singre microscopic field. The positionof each neuron in the fierd was marked on a ruled sheet of paper identicalin formar to the grid within the eyepiece. In this way the invesiigator couldbe certain which neurons were tabulated thereby preventing oversight orduplication.

Two types of glia fulfilling a standard criterion were counted: astrocy.teswith large, clear, brue-stained nuclei and origodendrocytes with smailer,deeply stained, blue nuclei. visual differentiation of astrogliar nuclei fromthose of small neurons is a frequently cited problerq in neuiocytologic workof this type. Previous studies attest to the effectiveness of creil echt violetin distinguishing between these ceil types (6, g). The grial counts wererecorded on the same sheets as were the neurons. To determine theneuronal:glial ratios, the counts of the astrocy.tes and oligodendrocyres werepooled from each section to provide a singre gJiat count. In addition,neuronal:astrocytic and neuronal:oligodendrocytic ratios were calculated.

shrinkage factors for frozen sections versus celloidin sections have beenconsidered in previous studies where we rearned that our experimentaldifferences between groups were the same whether we used ceiloidin orfrozen sections (3).

RESULTSTo test whether the Einstein brain diflered significantly from the popuration

from which the l l control brains were sampled, the mean and the standarddeviation for the sample were taken as estimates of the population parameters1t and o. Then, the deviation of the neuronal:glial ratio for Einsrern's brainfrom the mean neuronal:glial ratio for the sample was computed in standarddeviation units. This score was referred to a Student's t distribution withnine degrees of freedom, because two degrees of freedom were lost inestimation, one for the mean and one for the standard deviation. Theresults of this analysis suggested that in left area 39, the neuronal;gjial ratiofor the Einstein brain was significantly smailer than the mean for thecontrol population (t = 2.62, df g, p < 0.05, two_tailed). Einstein,s braindid not.differ significantry in the neuronar:grial ratio from the contrors inany of the other three areas studied (see Table I ).

Neither the neuronar:astrocyic nor the neuronar:oligodendrocy,tic ratiosby themselves were significantry different in any oi tte ur.", ,,ud,.d,comparing Einstein's brain with the contror brains. It was necessary to poolall glial cells counted to attain statisticaily significant differcnces, but thedata indicated that one gliar celr type arone *as not responsible fbr thedifference noted.

i l t .rii

4i:,

202 DIA]VIOND ET AL.

TABLE I

Neuron:Glial Ratios belween Einstein's Brain and Those l iom l l lVtales(47 to g0 years of Age)

RcgionN:Gi

( l I males)N:G;"

SD Einsrein VoL

Left area 9Right area 9Left area 39Right area 39

1.8491.7 54t.9362.026

0.6610.7550.3120.588

L04l, l6l . t20.92

775l73

t20

NSNS

0.05NS

' ln every area Einstein had a smauer N:G rat io, but by comparing one brain rv i th t l havingrelatively rarge SDs, the resurts showed only one area to be significantry different.

DISCUSSION

we studied the prefrontar and inferior parietal association areas ofEinstein's brain because such areas are known to be concerned with"higher" neural functions. These regions do not directly receive primarysensory information, but rather, as their name implies, ,.associate,,

or.analyze inputs from other brain regions. The associaiion-cortices are thelast domains of the cortex to myerinate, indicating their comparatruety ratedevelopment. It is not possible at present to identify with a t igi oegree ofspecificity the independent functions of these ,on"r. Charactenzing themodes of function of the corticar association regions may prove to be oneof the most elusive of all neurobiological tasks.

considering the fact that the tissue blocks were already embedded inc_elloidin when they became avairable for histologicar study (thereby makingGolgi or other more revearing studies impossibre), we decided that differentialcell counts constituted a potentiaily meaningfur measure of the functionalstatus of the brain. Not only is the cerebral cortex rich in its distribution ofnerve cell bodies, but grial ceil types arso constitute a large fraction of themammarian cerebral cortex. Bass e/ at. (2) reported thlt neu.onar:griarratios decrease as the phylogenetic scale is ascended from mouse to man.on the other hand, Rocker et at. (r6) demonstrated remarkabre consistencvin the absolute number of nerve ceils in cortical strips from pi"r ,".r*"'i"white matter, regardress of the mammalian species or cortical thickness.Such uniformity in number was found, for instance in the motor cortex(area 4) and in the somatosensory cortex (area 3b), arthough not in thevisual cortex (area l7) which has about two-and-one-half t imes as manyneurons as other cortical areas.

The thicker cortices of large mammals seems to be primarily a functionof large nerve cell bodies, more extensive dendrit ic and axonal systems, andconcomitantly, more numerous glial cells. Furthermore, environmentar

EINSTEIN'S BRAIN CELLS 203

ennchment and other augmented neural inputs in the rat increase all theseneuronal measures of enhanced cell activi iy together *i,r., ""

rncrease inthe number of gJial cel ls ( l , 3, 5,7, g, I l ) .An increase in the number of gJial cets without a significant increase inthe neuronal population suggests a response by gJial cells to g..ot.. neuronalmetabolic need. Alr these data suggest that neuronar;gJiar ratios in serectedregions of Einstein's brain might reflect the enhanced use of this tissue inthe expression of his unusuar conceptuar powers in compariso, *i,;;;;,r;;

brains.The rationaie for choosing the prefrontal and infraparietal regions wasbased on the specurations of several investigators. co-p-oi*" anatomicarstudies indicate that

-the parietal lobe expand, progr.rriuely to crowd themotor, auditory, and visual cortices forward, downward, and backward,respectively' studies,of.endocasts by von Bonin (17) comparing panetaland frontal lobes led him to conclude that it was this exiansion of theparietar lobe which was most characteristic of the human brain. Accordingto Passingham (15),-on the other hand, the prefrontal cortex is thought tosubserve in unique fashion those activities and quarities which distinguishman from other mammals and primates. The anterior portion-ortire frontallobe appears to be engaged in the temporar organization oit"iuuio., ..g.,the planning and estabrishment of behivioral itrategies (t3). From tesionstudies in animals and human beings, it has been sho*n thaitheprefrontal

cortex is involved in mechanisms of attention, recent memory, capacity forabstracting and categorizing information, and the formuration and init iationof actions- The parietal robe has been associated with tr,e in-t-egration otvisual, auditory and tacti le modalit it ies and with problems of self-lwareness,imagery, memory, and, attention (r4). Lesions in the inferior parietal region(area 39), especia,y of the dominant side, resurt in inabirity io read wordsor letters, and in gross impairment in writ ing, spering, aJ catcutation[(12), for recent review see (9)].

one mathematician with a resion in area 39 found it diff icult to draw orwrite formurae and courd not use a slide rure. Howev.., u, nigii he couldvisualize the correct construction of the formulae (3). A mathematician atthe University of California, Berkeley, calvin Moore, ,,ur.o ,r,u, tr. d.u"lop,a feeling of reality for abstract concepts. They exist in his brain and can bemanipulated like real objects. It is the interpray of these obj; *r,i.h -uu

contribute to mathematicar insight. It has also been .eporteJ ,n"i"i""in;educated individuar, resions in the inferior parietal robule of the dominanthemisphere result in the loss of versati l i ty oi i-ug.ry and the capabil ify forcomplex th inking (3).The possibre relationship of these phenomena to Einstein,s inte'ectualgifts served as a guide for the selection of our tissue samples. It thereforeseemed conceivable that area g of the prefrontal cortex and/or area 3g of

.!,,

20.1

the inlerior panetal cortex on rhe left and/or nght sides might be charecterized

by smaller than normal neuronal:g-l ial ratlos'

Our data suggest that the neuronal:glial ratio in area 39 of the left

hemisphere in glnsteinls brain is significantly lower than that of the control

subjects, or of the other regions in which measurements were made (e'g' '

area3g,r ight :areac,tef tanar ight) .Mentalact iv i t iesascr ibedtoarea39fit many of the comments that Einstein himself made about his conceptual

processes.

REFERENCES

l. ALTMAN, J., AND G' D' Drs' 1964- Autoradiographic examination of the effects of

enriched .nu,ron..nion tt. .ut" of g.lial multiplication in the adult rat brain' Nalttre

(London) 204: I l6 l - l 163'

2.BAss.N.H.,A.HESS,A'Pope,eupC'THeLsettueR'1971'Quant i tat ivec} ' toarchi tectonicdistr ibut ionoin. , ,on, ' -gt iaandDNAinratcerebralcortex.J.Comp,Neurol ' |43:

. 48 l -490.

3. CntrcHr-rv, MncDonnlo' 19'14' The Pariera! Lobe' Hafner' New York'

4. DIAMoND, M. C., F. f-o*' ff' RHoDES' B' LtNoNrn' M' R' RosENzwEtc' D KRECH'

nNo E. L. SrNNerr' t9e6- Increases in corrical deprh and glial numbers in rats subjected

to enriched envrronment' J' Comp' Neurol' 128: ll ' l-126'

5. Dtnt'tor'ro, M. C. 197;- Anatomicai brain changes induced by environment- Pages 215-

241 in J. tlcCructr ;; i' PErRlNovtcH' Eds'' Knowing' Thinking' and Bclieving'

Plenum' New York '

6. Drnvoxo, M. C., R. E. JoHNsoN, eno M' W' GoLD' 1977' Changes in neuron and gl ia

number in the young' adulr' and aging rar occipital cortex' Behav' Biol' 20:4O9-418'

7. DIAMoND, tn l . C., nNo": ' R' Couot ' Jn ' l98l ' l

t : " t : l for the potenual .of

the agtng

cortex. Pages 43-58 in S- J' ENNA' et al ' Eds' Brain Neurotransmilters and Receplors

in '4gittg and Age-Re'lated Disorclers' Aging' Vol' l7' Raven Press' New York'

8. Dtatv loNo, fU. C., o,u" : ' -n ' Co*no* ' fn ' tgS+' Morphological measurments in the aging

rat cerebral .o.t.*. Pug"' 43-55 in S' w' SCHEFF' Ed'' Aging and Recovery of Function

in thc Cantral Nervous Srslern' Plenum' New York'

g.ErDEl-BERc'D.,ANDO'f t ' 'Cot^uunon' lg84lnfer iorpar ieral lobule 'Art 'h 'Neurol4l :

843-8 5 2.10. ElNsrEIN, A

York.t l . KuHLENKAMPF, H. 1952' Das Verhal ten der

enrnarkes der weissen Maus unter dem

Emtu' ick. I l6: 304-3 I 2.

12. LYNCH, J. C. 1980. The funct ional organrzatronof posterior parietal association cortex'

I954. Page 25 in E. SEELIG, el al' ' Eds-' Itleas and Opinions Bonanza' New

neuroglia in den Vordenh6rnen des Riick-

Reiz physiologischer TZitigkeit' Zeil' Anat'

1980. The basic uni formrtY tn

Univ. of Chicago Press' Chicago

l l .

t4

Bchav. Bruin Sci. 3: 485-514.

MAsTERToN, R. B. ' AND M'-n ' Brerr-eY' lg74' Brain funct ion: changing ideas on the

ro|col 'scnsory,motoran<lassociat ioncortcxinbchavior.Annu.Rav.I 'syclnl .25:2.71-312'

^^ i i - - r narw^rk l r r r r l i rected at tent iJn and uni lateral ncglcct .MLSULAM. M. M. 1981. A cort ical network for d i recte '

, lnr t . Nttrr t t l . l ( ) : 309- '125'

15. PAsstN(; l lAM, R. E., AND G ETTLINcER IgT'1 A comparison of cort ical lunct ions tn man

and other primates. Inl' Ret'' Ntxrobiol 16:233-299'

16. RocKEL, A. J. , R. w. HIoRNS' AND T P' S' PowELL'

struclure of the neocortex' Brain 103:721-244'

17. VoN BoNtt ' t , C. 1963.The Etnl t t t ion of thc I Iunrun Bruin '

DIr\NtOND

I

1 l

I