AN EXAMINATION OF THE BREITINGER METHOD O F CRANIAL CAPACITY DETERMINATION T. D. STEWART Division of Physical Anthropology, 17. S. National Muaeum At the present time the determination of cranial capacity, despite the countless efforts at refinement, offers greater chances for error than most other measurements generally taken by craniometrists. If this measurement is worth taking at all, and probably no one will deny that it is, there should be available a standard method that is both simple in its procedure and reliable in its results. Until this point is reached there will continue to be a lack of confidence in the records. During the past year (’36) Breitinger published a method for which he claims both simplicity and reliability. He states, moreover, that numerous repetitions of the measurement by observers who used the method for the first time, showed almost as small an observational error [as his own], and therewith demonstrated that this technique, given personal demonstration-which is indeed indispensable for all methods of measurement-ensures the desired reliability and exactness in other hands too (p. 144). Regardless of the value of personal instruction in anthropometry, a continuity of this instruction is not entirely feasible, and hence I feel that it is desirable to subject this method to the severer test of being employed by one who has only the printed instructions to follow. That I have undertaken this project is due largely to the urging of Miss M. L. Tildesley of the “Comit6 de standardiAlso, i t must be zation de la technique anthropologique.” ’ I am indebted to Miss Tildeeley for supplying me with a translation of Breitinger ’8 paper. 111 AYBRICAN JOUENAL OF PEYSICAL ASTHROPOXLIGT, VOL. XXIII, NO. JGLY--SEPTEMBER, 1937 1 112 T. D. STEWART admitted, my willingness to do this work was influenced by my interest in cranial capacity as a result of having published recently (’34) on the general subject. GE-NXRAL PRINCIPLES It is evident from Breitinger’s paper that he has clearly recognized the general principles involved in the determination of cranial capacity. However, like others, he has not, in my opinion, placed sufficient emphasis upon them. Before outlining his method it may be well, therefore, to state briefly some of these principles. I believc. all will agree that the following a r e the most important: 1. Any standard technique for filling the skull must reproduce a certain density of the filling material (seed) as constantly as possible. 2. The density of seed in a skull of unknown capacity can only be inferred from previous experience with the same filling method on standard skulls (crtnes 6talons). 3. To obtain the capacity of a skull by use of a measuring cylinder the density of the seed in both cylinder and skull must be the same. 4. The density of seed resulting in a measuring cylinder when the latter is used in combination with a funnel depends chiefly on four factors: a) the size of the seed, b ) the type of funnel (slope of side and width of opening), c) the distance the seed fall, and d ) the diameter of the cylinder. Larger seed2 flowing a t a slower rate (narrower funnel opening) from a greater height and into a narrower cylinder all contribute to an increased density. METHOD The method advocated by Breitinger, using mustard seed, is a n adaptation of that of Mollison (’32) designed for the use of rape seed (‘Riibsamen’). Briefly, Breitinger determined the density given by his cylinder-funnel combination As compared with smaller eeed of the aame kind; for instance, large and small mustard seed. CRANIAL CAPACITY DETERMINATION 113 (principle 4, above) and worked out a technique for filling the skull that gave this same density (principle 3) fairly constantly (principle 1). The preliminary testing was done, of necessity, on a standard skull (principle 2). The cylinder in use had an interna.1 height of 50 cm., and a n internal diameter of 73 mm. The funnel had a slope of 60” (“sein Vertikalriss ein gleichseitiges Dreieck darstellt”) and a 15 mm. opening. With ‘top medical quality’ mustard seed this combination produced a density of 767 gm./1000 cc. (0.767). To reproduce in the skull this density of seed Breitinger developed the following procedure (literal translation, p. 147) : 1. Bed skull on pad, frontal obliquely downward. 2. Pour in the mustard seed through funnel ; the seed level should be kept u p to one-third the height of funnel; through gentle tilting of the skull on its bed the seed which fill the funnel to this level when they have stopped running will all flow into the skull. 3. Shake the skull with the hands laid on flat, frontal obliquely down; rate: four movements back and forth per second; duration: a) 15 seconds holding skull by frontal and occipital; b) 15 seconds holding skull by the sides. 4. Pour more seed in slowly straight from jar ; at the same time tilt skull on the bed in all directions with the left hand. 5. Distribute the masses of seed by probing six times with the wooden stick: once each slanting forward and to the right, sideways and to the right, backward and to the right, and similarly for the left. 6. Again slowly fill up from the j a r with a simultaneous tilting of skull on its bed. 7. Distribute seed for the last time with the thumb; fill in to the edge of the foramen magnum. 8. Empty contents of skull into funnel which has been set in position on cylinder ; fill funnel completely. 9. Let it run steadily into cylinder. ‘This stick is not described, but is assumed to be the same as tlint figured by Mollison; namely, 20 mm. in diameter mid rounded a t the ends. 114 T. D. STEWART More details are given in the text (pp. 142-143), but the above will suffice to orient the reader. Circumstances warranted the following changes in this procedure for the purpose of the present tests : a) Instead of placing the skull on a pad and filling by means of the Mollison funnel (steps 1, 2 and 4), the skull was held at a n angle as directed under the seedcontainer of the Goldstein machine (Stewart, '34, plates 1 and 2). This required that the end point of the first stage of filling be estimated. b) Not having a Mollison brass cylinder and funnel, the current apparatus of the laboratory was used for steps 8 and 9 ; it is the same as was used in 1934 (to the description on p. 343 may be added the fact that the funnel has a slope of 90"). I believe it will appear that neither of these alterations have seriously impared the effectiveness of the test. TESTING The logical beginning for a test of this method is to fill a series of standard skulls with mustard seed by the procedure just outlined, in order to see how nearly a density of 767 gm./ 1000 cc. is reproduced. The results of two such trials are shown in table 1. It will be seen that with the seed available in this laboratory I obtained an average density of 764 gm./1000 cc. in the skulls by Breitinger's method of filling. The range is 756 to 774, or a difference of 18, which is greater than the 10 recorded by Breitinger for fifty-three trials on the same skull. The cylinder also gives a variable density, owing chiefly to the fact that readings were made to the nearest 5 cc.' Because the cylinder-funnel combination gave a density greater than that achieved in the skull (769: 764), the average for the cylinder readings is lower than the true capacity of the series. The most important conclusion to be derived from this is that the mustard seed in this laboratory differs from that nsed by Breitinger. I n support of this conclusion I may state ' 1393 would be read 1395; 1398 = 1400; 1397 = 1395, etc. 115 CRANIAL CAPACITY DETERMINATION that subsequently I deliberately tried to create a greater density of seed in the skulls by more vigorous shaking, more forceful poking with the stick and thumb, and even by slapping with the hand, but all to little effect. I decided next to see how much effect size of seed had upon the results. For the purpose of cleaning the seed this laboratory has long used a sieve made of galvanized wire, 16 mesh TABLE 1 First results with the Breitinger method TRIAL SKULL WAl'EB NO. I- E- gb -. __I-- cc. 244080' 341791 244063 255051 225035 244055 244076 244074 244067 341794 Average I 1345 cc. I 1340 nm. 822 917 986 1023 1042 1049 1063 1077 1134 1166 1028 .- O h .-34 -"23 6% .e Clo ;b ~- ~ 1065' 1195 1285 1330 1360 1365 1380 1400 1475 1075 1210 1290 1330 1365 1370 1390 1410 1475 - -__ .e * 34 2 .E - TRIAL 2 I nm./ 000 CC. 772 767 767 769 766 768 770 769 769 765 768 Bm.1 1000 cc. 765 758 764 769 763 766 765 764 769 757 764 cc. 1065' 1190 1280 1330 1360 1365 1380 1420 1460 1530 -1340 - gm. 820 915 984 1023 1047 1050 1064 1092 1122 1172 1029 -gm./ -0 .-xmd 5; n -- nm./ , 0 0 0 cc. 1000 cc. 770 769 769 769 770 769 771 769 768 766 769 763 756 763 769 767 766 765 774 761 761 764 l F o r a description of these standard skulls see Stewart, '34, p. 345. Figures in this column rounded off to neareat 5 cc. t o the inch. Few seed, even when broken, go through this sieve, but all sand and dust are passed. By substituting a sieve having 12 mesh to the inch,5 I was able to remove quantities of small seed.s ' A 10-mesh copper screen was not as effective because the individual wires were thicker. 'Each seed i s somewhat irregular in shape and may or may not pass through the sieve according as different faces are presented. Hence, even after l011g shaking some few seed continue to pass through the sieve. 116 T. D. STEWART The next step was to adjust the cylinder-funnel combination so as to get a seed density agreeing with that for the skull given by the Breitinger method of filling. This is the reverse of Breitinger's problem. Thus, a new series of trials were run, using the larger seed and varying the funnel.' Instead of the metal funnel previously used, which now gave a density of about 775 gm./1000 cc., one made of cardboard was substituted. The new funnel differed from the old chiefly in having a 60" slope and in lacking a spout (said not to affect the flow of seed-Breitinger, p. 142). The round opening was varied either by further cutting away the cardboard or by applying adhesive tape. I n using this funnel the seed were emptied into it all at once from a container which had received them from the skull. The results of eight trials are shown in table 2. First of all it should be noted that the average cylinder reading approaches the average water volume as the densities in skull and cylinder likewise approach one another (principle 3). The best results were obtained in the final trials (7 and 8) where the average volumes are 2 cc. apart and the densities only 1 unit apart. Examination of the individual readings shows that in the last two trials the largest difference in capacity was 11 cc. On the other hand, the density of seed in the skull, the uniformity of which is the aim of the procedure, varied during eight trials on each of the ten different skulls between 760 and 778, or 18 units. However, except for two skulls (244080 and 225035) the greatest range in density for any set of eight trials was 10 units ; the lowest range was 2 units for skull 244067. This would suggest that Breitinger's findings, based for the most part on a s i n d e skull, a r e a little too optimistic as regards general applicability. ' Variations in the height of the cylinder are not very effective in changing the density of the seed. Varying the diameter of the cyliuder is of course out of the question. TABLE 2 ECght trials with the Breitinger method, w i n g large seed and varying the funnel opening DENSITY IN CYLINDEB SKULL NO. DENSITY IN SKULL 1. Fuunel oneuing = 2 1 mm. om./ m./ -- ce . 1075 244080 1210 341791 1290 244063 1330 244051 1365 225035 1370 244055 1390 244076 1475 244067 1540 341794 2794461 -1605 Average 1365 1 0 0 0 ee. 1 0 0 0 ce. 765 771 773 774 772 774 772 772 772 765 771 761 768 7 70 776 774 771 765 770 7 74 -762769 >YLINDER VOLUME 2 Same -__fuunel as in 1 ec. I Same Same Same Same Same Same Same Same Same Same _ Same Same Same Same Same Same ’ Same Same 1 Same Same Same Same Same Same 1 Same Same I Same Same Same Same i Same ~Average 1 Same I i 767 778 766 767 764 770 764 775 765 778 765 769 766 768 766 769 763 773 764.766 _ 765 771 .~ n g = 2 1 mm. 1082 831 768 773 927 767 1208 766 990 760 1302 767 1025 759 1350 771 1054 763 1381 772 1059 765 1385 773 1064 765 1390 765 1135 766 1482 769 1193 1557 766 775 1608 -__ 1232 766 768 764 770 ing -_ = 20 -mm. iel _ _opt . __ 826 770 768 1201 774 930 769 986 1289 765 764 768 1025 1335 771 1044 1360 768 765 1374 769 1056 771 768 1065 1387 766 769 1134 1474 769 769 1185 1540 769 1599 1230 - 769 766 ___ -I. - 1363 I 1048 769 I 768 _ _ ~ ~ nm. 1 0 0 0 cc. gm’/ 774 829 772 932 774 998 1026 773 1056 772 1064 I 774 1063 772 1135 772 1184 769 1228 767 1052 772 1071 1208 1290 1328 1368 1375 1376 1471 1540 1601 1363 ing= 22 mm. Same Same Same Same Same Same Same Same Same ~Same Averaee WEIGHT O F SEED 1 771 770 774 771 774 I 777 765 769 iG9 765 570 I 1 4. 1085 1202 1300 1349 1380 1380 1391 1485 1560 1609 1374 921 993 1032 1055 1056 1067 1137 766 764 765 764 765 767 766 765 825 928 989 1022 1054 1052 1072 768 768 761 762 765 766 766 6 1074 1208 1300 1342 1378 1374 1399 1483 1553 1611 1372 -_ - __ - i72 761 770 776 773 771 768 771 775 -766 770 - 767 7G 7 767 i68 772 768 771 769 774 - 769 . 7G9 __ inelas ii - 8 Same Same 769 1069 774 Same 766 760 Same 1199 919 Same 766 1290 Same i66 988 Same Same 1336 1025 767 i71 Same 769 Same 1365 769 1050 Same Same 13i2 768 769 1054 Same 769 76.5 Same 1383 1064 Same 769 7G9 Same 1475 1134 Same 770 571 Same 1542 1187 Same Same 766 1599 1229 I 769 -4verage I Same 1363 1 1048 1 7 6 9 1 ’ i G S .~. _ _ . ISubstituted for 244074 which mas no longer available. It is the skull of a male Alaskan Eskimo and has a cranial index of 75.3. Three trials 1vit.h water gave the following results: 1600, 1605, 1605 cc. 117 118 T. D. STEWART DISCUSSION This examination of the Breitinger method would seem to be a severe test because it duplicates in a measure the circumstances under which any standard procedure must operate. However, I believe the results of these trials will convince almost every one that, with the minor changes indicated, this method will yield satisfactory results even without personal instruction. One necessary modification involves the density of the seed. Breitinger’s report implies, although he may not have intended it, that a density of 767 g m . / l O O O cc. should be adopted as a part of the procedure. As here shown, it is not likely that every laboratory will have the same quality of seed, and therefore will not get this density figure. Lacking such standard seed, it would be foolish to standardize the density. The present results suggest, at least, that when the quality of the seed varies, the density in skull and cylinder varies equally. This remains to be proved definitely. However, if true, the Mollison cylinder and funnel, used with the Breitinger technique, should give reliable results regardless of seed quality. Rather than standardize the apparatus I would suggest that some such course as that followed in the present study be recommended. This would permit the adaptation of existing apparatus and avoid undue expense to the laboratories concerned. I n this coniiection there is something to be said in favor of a cylinder both taller and narrower than that described by Mollison ; namely, the greater accuracy in reading. Also, since a tall slender cylinder increases the density, this can be compensated for by a larger funnel opening, which in turn speeds the flow of the seed, thus saving time. Turning to the details of the skull-filling procedure, I would like to state some impressions not mentioned in connection with the testing. I n the course of handling the standard skulls during this study I came to feel that the variations from true capacity were due in part to variations in skull morphology. F o r instance, in one skull the foramen magiium was CRANIAL CAPACITY DETERMINATION 119 so small that the stick could not be introduced at much of a n angle; in another the occipital region was so formed that the last stage of filling progressed with unusual rapidity. Thus perhaps the range of trial capacities for the individnal standard skulls reflects somewhat the varying morphology of these skulls. This is a good reason for making tests 011 more than one standard skull. Again, at two stages in the procedure the directions were not clear enough. It is stated that the initial filling should continue until the skull is four-fifths full (p. 143). This point is difficult to judge and I feel that it considerably influences packing, at least in the extremes of skull size. Also, in connection with probing with the stick, the position of the skull is not indicated. I believe that packing is more effective at this stage if the skull is held with the frontal directed downward. Thus far we have not been concerned with the last two sections of Breitinger ' s paper ; these also deserve some coniment. After developing his method, Breitinger has sought to demonstrate its efficiency by showing how accurately the capacity of skulls of unknown volume can be determined. Taking ten Melanesian and ten Bavarian skulls, he determines their volume by this method and then also calculates the same from the weight of the seed, using a density factor. Although not stated, the density factor used in this calculation is presumably that which the method produces, namely, 767 p./ 1000 cc. It is obvious that this figure is the same as that given by the cylinder-funnel combination, and therefore it is far from gratifying that the same volumes should result; indeed it would be peculiar if they differed. That a particular density of seed is obtained i n a skull of unknown capacity can only be inferred from experience with standard skulls ; it is not proved. I would like also to discuss Breitinger 's statement (p. 146) that the capacity of the second standard skull by his method was the same before and after water-proofing; in other words, that he failed to substantiate my finding of a change in capacity under these circumstances (Stewart, '34, pp. 348-351). 120 T. D. STEWART It will be recalled that I was attempting to find out whether the condition of the internal surface of the skull affected seed packing. To this end I tested skulls of unknown volume by two different methods before and after coating the interior surfaces with paraffin. The main difference between the two methods was in the degree of seed packing (density). One method, which I designated ‘minimum packing,’ gave a density of about 747 gm./lOOO cc. ; * whereas the other method, called ‘maximum packing,’ gave a density of about 798 gm./1000 cc. Both methods were felt to be reliable for standard skulls within a possible error of 15 cc. The demonstrated effect of altered internal surface upon the results by the ‘maximum packing’ method was so striking that it was presented for what i t was worth. I have not been able to follow it up further. That Breitinger’s result does not disprove my finding is due probably to the fact that the seed density produced by his method (767 gm./lOOO cc.) is scarcely beyond the ‘minimum packing’ class, for which no variation in capacity due to altered internal surface was claimed by me. SUMMARY 1. Breitinger has reported a standard procedure for filling the skull with mustard seed for which he claims uniform results, even in the hands of others who have been instructed by him. 2. On the suggestion of Miss M. L. Tildesley of the ‘‘ComitCI de standardization de la technique anthropologique’ ’ the author has tested this method, using only the printed instructions and the available apparatus. 3. It was found that the density of the seed depends in part upon their quality and therefore the density should not be standardized. 4. Using a convenient seed density, results essentially comparable with those of Breitinger were obtained. ’A typographical error makes this 757 in the original, page 349. CRANIAL CAPACITY D E T E R M I N A T I O N 121 5. The findings are discussed and four general principles essential to the correct determination of cranial capacity are emphasized. LITERATURE CITED BREITINGER, EMIL 1936 Zur Messung der Schadelkapazitiit mit Senfkornern. Anthrop. A m . , XIII, 140-148. MOLLISON, TR. 1932 Hohlraummessung und Volumenbestimnung. Anthrop. Am., VIII, 290-294. T. D. 1934 Cranial capacity studies. Am. J. Phys. Anthrop., XVIII, STEWART, 33 7-361.