DEPARTMENT OF MEDICAL HISTORY

Department of medical history

The exceptional brain of Albert Einstein

Sandra F Witelson, Debra L Kigar, Thomas Harvey In recent decades, there have been major advances in neuroscience at the behavioural and neural levels, but the long-standing issue of the neurobiological basis of variation in intelligence remains unresolved.1 Around the turn of the 20th century, much attention was focused on anatomical correlates of intelligence through detailed necropsy case studies of the brains of outstanding people, such as mathematician Karl F Gauss or physician William Osler.2,3 By 1907, Spitzka4 had published an extensive monograph that summarised 137 case reports of notable men and women such as Bach and Descartes, and also presented one of the first group studies of nine scholars. Weight of the brain and patterns of gyral convolutions were usually examined. This early work had several limitations. First, medical and cognitive status at the time of death were often not known. Second, normal comparison groups were not available, so that the results were mainly idiosyncratic observations. Quantitative measurement was usually limited to the weight of the whole brain, and even its relation to intelligence remained unresolved. For example, novelist Ivan Turgenev’s brain weighed 2012 g,4 whereas the brain of author Anatole France was half the value (1017 g).5 Third, work was based on the assumption that intelligence was a unitary homogeneous ability—even though different people varied greatly in their area of cognitive excellence. (According to current theories of intelligence, there are independent spheres or modules of cognitive ability.6) Last, the studies had no a priori hypotheses as to the relation between structure and psychological function, since there was little knowledge about the cortical localisation of cognitive function.7 After the horrific events of World War II, issues related to the neurobiological substrate of intelligence were considered with great caution, and research in this area dwindled. The development of computerised imaging technologies has made it possible to obtain quantitative measurements of brain anatomy in vivo with magnetic resonance scanning, and renewed attention has been directed to the investigation of structure-function relations in the general population. The studies have varied greatly in their methodology, and, although the results are inconsistent, they do point to a low, but statistically significant, positive correlation between brain
Lancet 1999; 353: 2149–53
Department of Psychiatry and Behavioural Neurosciences, McMaster University, Hamilton, Ontario, Canada (S F Witelson PhD, D L Kigar, T Harvey MD) Correspondence to: Dr Sandra F Witelson, Department of Psychiatry and Behavioural Neurosciences, Faculty of Health Sciences, McMaster University, HSC 3G53, 1200 Main Street West, Hamilton, Ontario L8N 3Z5, Canada (e-mail: witelson@mcmaster.ca)

volume and IQ scores.8 Further work is needed to reconcile these results with the inconsistent findings on brain weight in the earlier case reports. Brain volume and weight are not perfectly correlated, and imaging does not provide measures of brain weight.

The case of Albert Einstein
Resolving the neurobiological substrate of intelligence may be facilitated by the comparison of extreme cases with control groups within the framework of specific hypotheses. Albert Einstein is one of the intellectual giants of recorded history, and the preservation of his brain provides the possibility of an important case study. Since Einstein’s death, there has been no report of the gross anatomy of his brain. Here we present the first such study. Our investigation of Einstein’s brain was guided theoretically on the basis of current information of cortical localisation of cognitive functions. The generation and manipulation of three-dimensional spatial images and the mathematical representation of concepts would appear to be essential cognitive processes in the development of Einstein’s theory of relativity.9 Einstein’s own description of his scientific thinking was that “. . . words do not seem to play any role”, but there is “associative play” of “more or less clear images” of a “visual and muscular type”.10 Visuospatial cognition,11,12 mathematical ideation,11 and imagery of movement13 are mediated predominantly by right and left posterior parietal regions. We hypothesised that the parietal lobes in particular might show anatomical differences between Einstein’s brain and the brains of controls.

Preservation of Einstein’s brain
Einstein died from a ruptured aneurysm of the abdominal aorta in 1955 at the age of 76 years. His medical history has been well documented, and his biographies show that he was mentally adept to the end of his life.9 Within 7 hours of death, his brain was removed at necropsy, fresh weight was measured, perfusion of 10% formalin by injection into the internal carotid arteries was carried out, and the whole brain was then freely suspended in 10% formalin for fixation and subsequent study. No significant neuropathology was seen on examination (gross or microscopic). After fixation, caliper measurements were made directly from the brain; calibrated photographs were taken of all views of the whole brain and of the dissected hemispheres; the cerebral hemispheres were cut into approximately 240 blocks, each about 10 cm3; and the location of the blocks was recorded on photographs. The blocks were embedded in celloidin, and histological sections were made. 2149

THE LANCET • Vol 353 • June 19, 1999

DEPARTMENT OF MEDICAL HISTORY

psychiatric status (as judged by clinical history and medical assessments) and normal cognitive ability (as documented by research neuropsychological testing that included IQ assessment).18 In each case, informed consent with respect to testing and necropsy had been obtained. Mean Full Scale IQ score on the Wechsler Adult Intelligence Scale19 was 116 (SD 9). Quantitative measures of Einstein’s brain and this control group were compared; Einstein’s brain was also compared with a smaller age-matched subgroup (in the collection) of the 8 men aged 65 years or more (mean 68) for brain measures known to change with advancing age. Although women have smaller brains than men,20 for purposes of descriptive analysis of gyral morphology, Einstein’s brain was also compared with 56 female brains (the total number of female brains in the same collection).

Measurements
Direct caliper measurements were made both from Einstein’s brain and from the control brains. Other measurements were made from calibrated photographs. We measured baseline values for overall dimensions of the brain, including variables for Figure 1: Photographs taken in 1995 of five views of Einstein’s whole brain (meninges removed) which there are published data A, superior; B, left lateral; C, right lateral; D, inferior; E, midsagittal view of the left hemisphere. The arrow in (eg, weight, corpus callosum each hemisphere indicates the posterior ascending branch of the Sylvian fissure as it runs into (is confluent with) the postcentral sulcus (compare with figure 2). Consequently, there is no parietal operculum in either size21); measures involving hemisphere. Scale bar, 1 cm. parietal regions important for Although there is no record of his having made specific visuospatial cognition and mathematical thinking; and, arrangements for post-mortem study of his brain, for comparison, measures of frontal and temporal Einstein was sympathetic to the idea of his brain being regions. Statistically significant differences between Einstein and the control group were defined as those studied. As reported in The New York Times in 1951, he, measures at least 2 SDs from the control mean. along with other physicists, underwent electroencephalographic recordings for research purposes.14 He also “insisted that his brain should be used for research”.15 At Einstein’s parietal lobes the time of his death, the family requested a necropsy, Figure 1 shows the set of photographs taken in 1955 of which was done by pathologist Thomas Harvey, who the lateral, superior, inferior, and midsagittal views of took the initiative to remove the brain for scientific Einstein’s brain. The superior view (figure 1A) shows a study. Consent was given by Einstein’s elder son, Hans relatively spherical brain which is corroborated Albert Einstein,16 and by the executor of Einstein’s quantitatively (see below). Moderate atrophy is present estate, Prof Otto Nathan (ref 17, p 264). around the main fissures in the central regions in both hemispheres, to an extent common for a person in their Control brain specimens eighth decade.22 A unique morphological feature is The control group consisted of all the male specimens visible in the lateral surface of each hemisphere which available at the time (n=35) in the Witelson Normal otherwise shows usual anatomy (figure 1B, 1C)— Brain Collection based at McMaster University. The key namely, the posterior ascending branch of the Sylvian features of this collection are that the brains are from fissure is confluent with the postcentral sulcus. research volunteers with normal neurological and Consequently, there is no parietal operculum (the

2150

THE LANCET • Vol 353 • June 19, 1999

DEPARTMENT OF MEDICAL HISTORY

anterior part of the supramarginal gyrus), which normally develops between these two sulci during fetal life.23,24 This morphology found in each of Einstein’s hemispheres was not seen in any hemisphere of the 35 control male brains or of the 56 female brains, nor in any specimen documented in the published collections of post-mortem b r a i n s . 25,26 Figure 2 highlights this unique feature of Einstein’s brain in comparison with a typical control brain. Three main types of morphology of the Sylvian fissure and surrounding gyri have been described previously;27 in each type, the Sylvian fissure terminates or bifurcates behind the postcentral sulcus, and the parietal operculum is present. The tracing of the superimposed hemispheres of the control brain (figure 2, no 3) shows the typical right-left asymmetry in size and position of the Sylvian fissure and the parietal opercula.28 By contrast, the tracing of Einstein’s hemispheres (figure 2, no 6) shows the confluence of the posterior ascending branch of the Sylvian fissure and the postcentral sulcus in each hemisphere, the absence of the parietal opercula, and unusual symmetry between hemispheres of sulcal morphology in this region. Quantitative measurements of Figure 2: Lateral photographs and tracings of left (solid line) and right (dashed line) Einstein’s brain compared with the superimposed hemispheres of a typical control male brain (1, 2, 3) and the brain of male control group are shown in the Einstein (4, 5, 6) table, with relevant landmarks shown The photographs of the control brain show the parietal operculum in the left (stippled) and right in figure 3. Einstein’s brain was not (hatched) hemisphere, situated between the postcentral (PC) sulcus and the posterior ascending branch of the Sylvian fissure (SF), which originates terminates at statistically different from the control S. PC is the inferior end of PC at SF. The tracing ofat the point of bifurcation (q) and(3) shows the the superimposed hemispheres group on most measures. His brain asymmetry in position and size between the parietal opercula. The tracing of Einstein’s weight did not differ from the control hemispheres (6) highlights the confluence of PC and the posterior ascending branch of SF in each hemisphere, the absence of the parietal opercula, symmetry of the sulcal morphology group, from the age-matched between hemispheres. Comparison of the tracings and thethe relatively anterior position of the SF shows subgroup, or from published large age- bifurcation in Einstein, and the associated greater posterior parietal expanse, particularly in his left matched groups (table, measure 1). hemisphere compared with the control brain. Unfortunately, the volume of Einstein’s brain had not been obtained. Brain length, height, size of respectively (measure 24). Parietal regions typically the corpus callosum, and measures of the frontal and show anatomical asymmetry (table, control group, temporal lobes did not differ between Einstein and measures 19–2428). Einstein’s parietal lobes were controls. However, size of a specific gyral region in the symmetrical (compare with figure 2, no 6). This was due frontal operculum was different in Einstein’s brain from mainly to his left parietal lobe being larger than that of the control group. The possible association of this usual, resembling a right hemisphere in size and feature in relation to biographical accounts of Einstein’s morphology. atypical speech development17 will be reported elsewhere. Discussion By contrast, in the parietal lobes, there were striking quantitative differences. Each hemisphere of Einstein’s The gross anatomy of Einstein’s brain was within normal brain was 1 cm wider (15%) than that of the control limits with the exception of his parietal lobes. In each group (measure 5). Maximum width usually occurs hemisphere, morphology of the Sylvian fissure was across the end of the Sylvian fissure—the region of unique compared with 182 hemispheres from the 35 unique morphology in Einstein’s brain. The ratios of control male and 56 female brains: the posterior end of hemisphere width to height and of brain width to length the Sylvian fissure had a relatively anterior position, (measures 6 and 7) showed that in Einstein’s brain the associated with no parietal operculum. In this same parietal lobes were relatively wider and the brain more region, Einstein’s brain was 15% wider than controls. spherical (see figure 1A) than those in the control group. These two features suggest that, in Einstein’s brain, In Einstein’s brain, the parietal operculum was missing extensive development of the posterior parietal lobes in each hemisphere in contrast to control values of occurred early,24 in both longitudinal and breadth 6·1 cm2 and 3·6 cm2 in the left and right hemispheres, dimensions, thereby constraining the posterior expansion
1

THE LANCET • Vol 353 • June 19, 1999

2151

DEPARTMENT OF MEDICAL HISTORY

Einstein Left Age (yr) Height (cm) Overall brain measures 1 Brain weight, fresh (g) 2 Hemisphere weight, fixed (g) 3 Maximum height of hemisphere (cm)† 4 Length of hemisphere (OF) (cm) 5 Maximum width of hemisphere (cm)‡ 6 Ratio of width of hemisphere to height 7 Ratio of width of brain to length (mean OF) 8 Corpus callosum area (cm2 ) Frontal lobe (cm) 9 F-PreC 10 FC 11 FA 12 A-PreC1 13 PreC1-C1 Temporal lobe (cm) 14 TO 15 C1-C2 16 SS1 Parietal/occipital lobe (cm) 17 O-PC 18 OC 19 OB 20 OS 21 BS 22 C1-PC1 23 PC1-B 24 Parietal operculum (cm2 ) Right 76 176 1230 550·0 545·0 8·9 8·7 17·2 7·5§ 0·84§ 16·4 7·5§ 0·86§

Control group (mean, SD) Left Right 57 (11) 178 (8) 1400 (118)* 591·0 (46·0) 591·0 (48·0) 9·3 (0·6) 9·4 (0·6) 16·9 (0·6) 6·5 (0·5) 0·70 (0·07) 16·8 (0·6) 6·5 (0·5) 0·69 (0·07)

0·89§ 6·8 9·2 11·3 5·1 0·8 1·2 13·2 3·9 6·1 8·4 8·9 7·1 8·0 2·5 3·5§ 0§ 0§ 9·5 11·6 5·1 0·9 1·2 12·8 3·9 6·6 7·9 8·3 7·9 7·9 2·9 2·0 0§ 0§

0·77 (0·06) 7·0 (0·90)¶ 9·4 (0·7) 10·6 (0·6) 4·8 (0·4) 0·9 (0·4) 1·4 (0·5) 13·2 (0·5) 4·0 (0·3) 5·1 (1·1) 8·3 (0·8) 9·5 (0·6) 5·8 (0·9) 6·1 (1·1) 0·9 (1·1) 2·3 (0·6) 1·9 (1·0) 6·1 (3·4) 9·2 (0·8) 10·5 (0·6) 4·7 (0·4) 1·0 (0·4) 1·2 (0·4) 13·2 (0·5) 4·0 (0·3) 6·0 (0·9)** 8·4 (0·8) 9·3 (0·8) 7·2 (0·9)** 7·4 (1·0)** 2·4 (1·3)** 2·0 (0·6)** 1·1 (1·2)** 3·6 (2·1)**

Figure 3: Sketch of a typical brain showing the landmarks for defining the measurements shown in table
F, O, and T: frontal, occipital and temporal poles, respectively; PreC, C, PC: superior ends of the precentral, central and postcentral sulci, respectively; PreC1 , C1, PC 1: inferior ends of these sulci, respectively; A, point of origin of the anterior ascending branch of the Sylvian fissure (SF); B, point of bifurcation of the posterior SF; S, end of SF; S1, and C2, points of the shortest distance from S and C1, respectively, to the bottom of the temporal lobe; parietal operculum (stippled region), the anterior segment of the supramarginal gyrus which surrounds BS.

Control group consists of 35 men and an age-matched male subgroup (see text). *Our control mean of 1400 g is similar to values of other studies of large groups of white men of similar age range (30–70 years)—eg, mean fresh brain weight=1399 g, n=1433, mean age=53 years.20 For the age-matched subgroup, mean (SD) fresh brain weight was 1386 g (149). In a large study, mean fresh brain weight for a 70–80 year age group was 1342 g, n=253.20 †Maximum height usually occurs near the plane of point C (figure 3). ‡Maximum width of each hemisphere occurs over the end of SF (figure 3). §Statistically different (2 SDs from the control group) or reflect unique morphology. ¶Callosal area is larger in non-right-handers and decreases with advancing age.21 There is evidence to suggest that Einstein was not consistently right-handed.37 Einstein’s callosal area of 6·8 cm2 tended to be larger than his predicted value (5·9 cm2) when hand preference and age were taken into account.21 **Statistically significant right-left anatomical asymmetry within the control group (compare ref 28) (p