NEUROSCIENCE

ELSEVIER Neuroscience Letters 210 (1996) 161-164

LETTERS
Alterations in cortical thickness and neuronal density
in the frontal cortex of Albert Einstein

B r i t t A n d e r s o n a,b,*, T h o m a s H a r v e y
aDepartment of Neurology, University of Alabama at Birmingham, Birmingham, AL 35294-0007, USA
bNeur6,1ogy Service, Birmingham Department of Veterans Affairs Medical Center, Birmingham, AL, USA

Received 11 March 1996; revised version received 26 April 1996; accepted 26 April 1996

Abstract

Neuronal density, neuron size, and the number of neurons under 1 mm 2 of cerebral cortical surface area were measured in the right
pre-frontal cortex of Albert Einstein and five elderly control subjects. Measurement of neuronal density used the optical dissector
technique on celloidin-embedded cresyl violet-stained sections. The neurons counted provided a systematic random sample for the
measurement of cell body cross-sectional area. Einstein's cortex did not differ from the control subjects in the number of neurons under
1 mm 2 of cerebral cortex or in mean neuronal size. Because Einstein's cortex was thinner than the controls he had a greater neuronal
density.

Keywords: Intelligence; Brain; Einstein; Neurons; Cerebral cortex

Humanity is fascinated by the brains of geniuses [6,9,
16,19,21] and scientists have measured brain size and gyri
to explain the intellectual bounty of great men. While
such neuromorphologic investigations may now seem
quaint, at odds with politically popular beliefs of general
intellectual parity and the modularity of mental opera-
tions, these prejudices deserve re-examination. Intelli-
gence is not a mystical psychological process which con-
fers smartness in proportion to its presence [7], but a psy-
chometric index for the effects of neurobiological features
which affect the efficiency of neural operations [1,23].
The presence of such biological features is indicated by
the anatomical and physiological correlations between
brain and intelligence, .e.g. the brain volume-IQ correla-
tions [2,10,15,22,24-261.
Studying the brain of a genius can play a small and
titillating role in the quest to identify these neurobiologi-
cal features. Single case studies, while not definitive, may
point the way towards s.olutions and have a long history in
cognitive research [4]. These considerations prompted us
to undertake precise c~uantitative measurements of the
cerebral cortex of Albert Einstein, our most renowned
* Corresponding author. Tel.: +1 205 9349691; fax: +1 205 9334465;
e-mail: brittuab@aol.com.
genius. Using the optical dissector technique neuronal
density, the number of neurons under 1 mm 2 of cortex,
and the cross-sectional area of a random selection of cor-
tical neurons were measured and compared to five elderly
men.
Einstein died in 1955, at age 76 years, of a ruptured
abdominal aortic aneurysm. An autopsy was performed
7 h postmortem. The brain was fixed in formalin for sev-
eral months. After fixation the cerebral cortex was di-
vided into numbered blocks and embedded in celloidin.
The brain region from which the Einstein block was taken
is shown in the diagram in Fig. 1. The fact that the block
is no. 9 and probably includes brain tissue from Brod-
mann area 9 is coincidental. This diagram was prepared at
the time of the original processing of Einstein's tissue for
celloidin embedding. The celloidin blocks were kept in
alcohol until use.
The control subjects were all men, ages 63, 63, 64, 71,
and 79 years. All died non-neurological deaths and rou-
tine neuropathologic examinations were unremarkable.
The area of cortex studied for each subject was right
middle frontal gyrus midway between the frontal tip and
the central sulcus and corresponded topographically with
the available area of Einstein's prefrontal cortex. Similar
to Einstein's tissue, all control blocks were immersion

0304-3940/96/SI2.00 © 1996 Elsevier Science Ireland Ltd. All rights reserved

PII: S0304-3940(96) 12693-6
162 B. Anderson, T. Harvey / Neuroscience Letters 210 (1996) 161-164
Fig. 1. Schematic prepared at the time that Einstein's cerebral cortex
was originally dissected. The cerebral cortical block used for the cur-
rent study was no. 9.
fixed in formalin, followed by celloidin embedding and
maintenance of the celloidin blocks in 80% ethanol. Most
of the control tissue was kept in 80% ethanol for only
weeks whereas Einstein's tissue had been maintained in
this way for decades.
Sections were cut at a thickness of 40ktm using a
sliding microtome and stained with cresyl violet. Under
microscopic examination and using the Neurolucida com-
puter program (MicroBrightField, Colchester, VT) four
measurements of cortical thickness were made on a single
section at 200 ktm intervals (see Fig. 2).
With a 100x objective each mark, as illustrated in Fig.
2, was sequentially moved to. A grid of 20 × 20/,tm was
overlaid and neurons counted while focusing down
through the section. Neurons in focus at the start of the
section were not counted, nor were any neurons counted
that touched the lower and left grid exclusion lines. Neu-
rons were counted when their cellular and nuclear mem-
branes came into sharpest focus. A nucleolus was not
required for a neuron to be counted. Neurons were differ-
entiated from glia based on size and staining patterns. The
number of neurons per dissector was recorded as was the
depth of the dissector. Most dissectors contained zero or
one neuron. The m a x i m u m was four. For all brains except
one it was necessary to make four passes through the
cortex to count the 100 neurons that was the study goal.
In one brain, because of a thicker cortex, a sufficient
number of neurons was achieved with only three passes
through the cortex. This approach is basically the optical
dissector [8,12,20].
While these cells were being counted the cell tracing
software of Neurolucida was used to trace each cell body
area at the cell's greatest diameter to provide the cross-
sectional areas of a systematically random selection of
neurons.
Neuron density was computed by dividing the number
of neurons counted by the area of the grid square multi-
plied by the total dissector depth (measured from the mi-
croscopic section) for all dissectors on each pass through
the cortex. The reliability of these estimates was high
(>0.95 by Cronbach's a). To calculate the number of neu-
rons under 1 mm 2 of cerebral cortex the neuron density
was multiplied by the cortical depth.
The neuron density for each of the passes through
Einstein's cerebral cortex was compared to similar meas-
ures for the control group by the M a n n - W h i t n e y U-test.
The same test was used to compare the estimates of the
number of neurons under 1 mm 2 of cortex and the size of
the cerebral cortical neurons.
A non-parametric test was used because the normality
of the data could not be established with this small sample
size. The reason for using multiple measurements from
Einstein's and the controls' tissues as individual 'sam-
ples' was to strengthen the conclusion that Einstein's
greater results for some measures were not due to chance.
With only six specimens the odds that Einstein would
have had, for example, a greater measured neuronal
density was l/6 by chance alone. However, if Einstein's
neuronal density was truly the same as the control popu-
lation, the odds that repeated assessments of neuronal
density would have been consistently greater than the
controls should have grown smaller with each repetition.
By using each measure as an individual 'sample' and ap-
plying the M a n n - W h i t n e y U-test it was possible to
quantitate the odds for such results.
Einstein's cortex was thinner than the controls and
more densely populated with neurons. The mean distance
through the right frontal cortex of Einstein was 2 1 3 7 / t m
and for the control population it was 2659/,tm (P = 0.002,
U = 0.0, n's 20/4). Einstein had more neurons per mm 3 of
B,
,I ji C.
" //
"~,,~ {I'i
LI I
Fig. 2. The approach for measuringthe anatomical variables. (A) On a
single microscopic section of cortex a flat area of the cortical ribbon,
excludinggyral depths, was selected. (B) Along this strip, and using the
Neurolucida Program, four roughly parallel transepts were drawn per-
pendicular to the cortical surface and separated by 200/.tM. (C) A ran-
dom number between 1 and 80 was determined and the first counting
zone (marked by a dot) was marked this many microns along the first
transept (a). Additional counting zones (additional dots) were then
marked every 80/~M along the vertical meridians (b). The marking
continuedfor all four meridianswith the distance between the last mark
of a meridian and the white matter carried over to the start of the next
meridian (therefore, in this example, c + d = 80/~M). The microscope
was moved sequentiallyfrom each counting zone to the next by virtue
of the Neurolucida program. Neuronal measurements and counts at
each countingzone followed procedures described in the text.
 B. Anderson, T. Harvey / Neuroscience Letters 210 (1996) 161-164
cortex with a mean for the four dissector paths through
the cortex of 46 995 neurons/mm 3 compared to the mean
for the control population of 3 4 9 6 2 neurons/mm 3
(P = 0.015, U = 8, n ' s 19/4 ).
Einstein's cortex did not differ in the size of the neu-
rons' somas or in the number of neurons contained within
a column of cortex 1 mm 2 at the surface. The mean cross-
sectional area of the neurons in Einstein's right frontal
cerebral cortex was 63 ~em2 (35 SD), and 68 ~ m 2 (46 SD)
for the control population (P = 0.85, U = 32,713, n's 501/
132). The mean number of neurons in a column of cortex
1 mm z at the cortical surface was 98 654 for Einstein and
87 364 for the controls (P = 0.26, U = 24, n's 19/4).
Studies performed on Einstein's brain tissue in the
early years after his death were only qualitative, and not
showing any important differences from normal subjects,
were not published. One prior study of Einstein's cerebral
tissue has been formally reported. Diamond et al. [5]
measured the ratio of neurons to glia in the frontal and
parietal cerebral cortex. Diamond and colleagues reported
a greater glia/neuron ratio for Einstein's tissue. Since they
reported neither the absolute numbers of neurons nor their
size, their results cannot be directly compared with this
study. The results for neuron density and neuron number
obtained from this inve,;tigation are in general agreement
with prior estimates [13,.27].
The differences between Einstein's tissue and the con-
trol tissue is primarily due to Einstein having the same
number of neurons packed into a thinner cortex. A g e or
dementia-related atrophy are not explanations for this
finding since Einstein's. cortex was compared to elderly
controls and because there is no history that Einstein suf-
fered from cognitive decline.
The thinner cortex is not due to technical factors. Both
Einstein's tissue and the controls' were embedded in cel-
loidin, sectioned, and stained similarly. As the brain tissue
is completely dehydrated in an alcohol-ether solution as
part of the celloidin embedding process, the subsequent
duration of the storage in 80% ethanol should not result in
greater tissue shrinkage with longer storage. A precise
measure of the size of ~the tissue block for Einstein prior
to its embedding was not available to quantitate tissue
shrinkage. In general, celloidin embedding is felt to result
in less tissue distortion than paraffin techniques, and
probably frozen sections that are subsequently dehy-
drated.
Could the difference in cortical packing density in any
way account for Einstein's superior intellectual skills? An
increase in neuronal density could translate to an overall
increase in the total number of cortical neurons if Ein-
stein's brain volume were equal to or greater than the
control cases. Unfortunately, it is not currently possible to
provide an accurate measurement of total cerebral cortical
volume or surface area for Einstein. This measure is nec-
essary to translate a neuronal density measure into an
absolute neuronal number measure. While it is not possi-
163
ble to provide a measure o f total cerebral cortical volume
the available data does not suggest that Einstein's cerebral
cortical volume was greater than controls. Einstein's brain
weight was 1230 g, which is within the average range, but
below the mean, for men of his age [11,14]. In addition,
the length of the hemispheres (17.2 cm left/16.4 cm right)
and the width of the hemispheres (7.5 cm left/7.5 cm
right) are not significantly different from 'eminent' and
non-eminent men born in the 1800s [21].
An alternative consideration would be that an increase
in neuronal density might be advantageous by decreasing
interneuronal conduction time. Interneuronal conduction
time has been posited as a limiting feature of the devel-
opment of brain size and cortical connectivity [ 17,18]. An
increase in neuronal density has been implied as a possi-
ble mechanism to explain how women and men may
achieve comparable I Q ' s [27] even though w o m e n ' s
brains are smaller when corrected for body size [3].
This work would not have been possible without the
superb technical assistance of Valisa Rutledge.
[1] Anderson, B., G explained, Med. Hypotheses, 45 (1995) 602-
604.
[2] Andreasen, N.C., Flaum, M., Swayze, I1, V., O'Leary, D.S., AI-
liger, R., Cohen, G., Ehrhardt, J. and Yuh, W.T.C., Intelligence
and brain structure in normal individuals, Am. J. Psychol., 150
(1993) 130-134.
[3] Ankney, C.D., Sex differences in relative brain size: the mismeas-
ure of women, too? Intelligence, 16 (1992) 329-336.
[4] Berker, E.A., Berker, A.H. and Smith, A., Translation of Broca's
1865 report, Arch. Neurol., 43 (1986) 1065-1072.
[5] Diamond, M.C., Scheibel, A.B., Murphy, G.M. and Harvey, T.,
On the brain of a scientist: Albert Einstein, Exp. Neurol., 88
(1985) 198-204.
[6] Donaldson, H.H., A study of the brains of three scholars: Gran-
ville Stanley Hall, Sir William Osier, Edward Sylvester Morse, J.
Comp. Neurol., 46 (1928) 1-95.
[7] Gould, S.J., The Mismeasure of Man, Norton, New York, 1981,
352 pp.
[8] Gundersen, H.J.G., Bendtsen, T.E., Korbo, L., Marcussen, N.,
Moiler, A., Nielsen, K., Nyengaard, JR, Pakkenberg, B.,
S0rensen, F.B., Vesterby, A. and West, M.J., Some new, simple
and efficient stereological methods and their use in pathological
research and diagnosis, APMIS, 96 (1988) 379-394.
[9] Hansemann, D., Ueber das gehirn von Hermann v. Helmholtz,
Zeit. Pyschol. Physiol., 8 (1899) 1-12.
[10] Harvey, I., Persaud, R., Ron, M.A., Baker, G. and Murray, R.M.,
Volumetric MR1 measurements in bipolars compared with
schizophrenics and healthy controls, Psychol. Med., 24 (1994)
689-699.
[11] Ho, K.-c., Roessmann, U., Straumfjord, J.V. and Monroe, G.,
Analysis of brain weight, !: adult brain weight in relation to sex,
race, and age, Arch. Pathol. Lab. Med., 104 (1980) 635-639.
[12] Mayhew, T.M., A review of recent advances in stereology for
quantifying neural structure, J. Neurocytol., 21 (1992) 313-328.
[13] Pakkenberg, B., Total nerve cell number in neocortex in chronic
schizophrenics and controls estimated using optical dissectors,
Biol. Psychol., 34 (1994) 768-772.
[14] Pakkenberg, H. and Voight, J., Brain weight of the Danes, Acta
Anat., 56 (1964) 297-307.
[15] Raz, N., Torres, l.J., Spencer, W.D., Millman, D., Baertschi, J.C.
and Sarpel, G., Neuroanatomical correlates of age-sensitive and
164 B. Anderson, T. Harvey/Neuroscience Letters 210 (1996) 161-164
age invariant cognitive abilities: an in vivo MRI investigation,
Intelligence, 17 (1993) 407-422.
[16] Riese, W., Brains of prominent people: history, facts and signifi-
cance, MCVQ, 2 (1966) 106-110.
[17] Ringo, J.L., Neuronal lnterconnection as a function of brain size
(review), Brain Behav. Evol., 38 (1991) 1 ~ .
[18] Ringo, J.L., Doty, R.W., Demeter, S. and Simard, P.Y., Time is of
the essence: a conjecture that hemispheric specialization arises
from interhemispheric conduction delay, Cereb. Cortex, 4 (1994)
331-343.
[19] Rodin, A.E. and Key, J.D., Osler's brain and related mental mat-
ters, South. Med. J., 83 (1990) 207-212.
[20] Royet, J., Stereology: a method for analyzing images, Prog. Neu-
robiol., 37 (1991) 433-474.
[21] Spitzka, E.A., A study of the brains of six eminent scientists and
scholars belonging to the American Anthropometric Society, to-
gether with a description of the skull of Professor E.D. Cope,
Trans. Am. Phil. Soc., 21 (1907) 175-308.
[22] Strauss, E., Wada, J. and Hunter, M., Callosal morphology and
performance on intelligence tests, J. Clin. Exp. Neuropsychol., 16
(1994) 79-83.
[23} Vernon, P.A., Biological Approaches to the Study of Human
Intelligence, Ablex, Norwood, 1993, 416 pp.
[24] Wickett, J.C., Vernon, P.A. and Lee, D.H., In vivo brain size,
head perimeter, and intelligence in a sample of healthy adult fe-
males, Pers. Indiv. Diff., 16 (1994) 831-838.
[25] Willerman, L., Schultz, R., Rutledge, J.N. and Bigler, E.D., In
vivo brain size and intelligence, Intelligence, 15 (1991) 223-228.
[26] Willerman, L., Schultz, R., Rutledge, J.N. and Bigler, E.D.,
Hemisphere size asymmetry predicts relative verbal and nonver-
bal intelligence differently in the sexes: an MRI study of struc-
ture-function relations, Intelligence, 16 (1992) 315-328.
[27] Witelson, S.F., Glezer, I.I. and Kigar, D.L., Women have greater
density of neurons in posterior temporal cortex, J. Neurosci., 15
(1995) 3418-3428.