Volume 32 1923 > Volume 32, No. 128 > Maori somatology. Racial averages. Part V, by Te Rangi Hiroa (P. H. Buck), p 189-199
MAORI SOMATOLOGY. RACIAL AVERAGES.
V. (Continued from Vol. XXXII., No. 1.)
Having concluded the measurements taken, the following table groups the averages of the various absolute measurements.
TABLE XXXIII.—AVERAGES OF ABSOLUTE MEASUREMENTS.
Averages, whilst for comparative purposes, useful in assisting us to visualise a race or a number of persons as concentrated into one individual, do not in themselves convey any idea of the various types that may go to the formation of those averages. In the above table it will be seen that the average indices such as cephalic, nasal, etc., occupy a medium position. This argues a fair amount of racial intermixture. Of the 421 individual cephalic indices that form the medium or mesaticephalic average index of 77·7, it has been shown in Table XII. that 63·4 per cent. belong to that class. The rest, however, are made up of 14·9 per cent. long-headed and 21·6 per cent. broad-headed. From this it will be seen that amongst the living (in this series) there is a higher percentage of broad heads than long heads. This is contrary to what has been accepted in the past. Previous data however has been supplied by the work of Scott, Turner and Flower, which was done on skulls. For comparative purposes, I have converted this series to similar terms by adding Broca's two units of difference in cephalic index by raising the upper limit of the long heads to 76·9 and the lower limit of the broad heads to 82.
TABLE XXXIV.—CRANIAL INDEX IN GROUPS (Converted).
Comparing the above table with Table XII., it will be seen that in the living, the broad heads exceed the long heads by 21·6 per cent. to 14·9 per cent., but for the crania of the same series the position is considerably reversed. The long heads now exceed the broad heads by 39·1 per cent. to 9·2 per cent. These anomalies should be removed by our authorities on physical anthropology. If Broca's two units of difference between the cephalic and cranial indices is accepted, why should we continue to use the same figures to mark the boundaries between the long, medium and broad classes of heads and skulls? Much confusion and extra work will be saved by altering the figures of either the one or the other. As Professor Roland Dixon1 has thrown new light on Maori physical types from data based on the measurements of Maori crania, I will adhere to the classification adopted in the above table whilst comparing with Dixon's results.- i
DISTRIBUTION OF CEPHALIC INDEX.
FIGURE 1.- 191
Plain line = Total Series., Broken line = North Auckland Series. Average, Total Series = 77·7., Average, North Auckland = 78·6.
In Figure I. the plain line shows a frequency curve for the cephalic index of the whole series. It will be noticed that the mode almost coincides with the mean of 77·7, and the curve is practically normal. This would convey the impression that the series is homogeneous. Such, however, is far from the case. In the Figure, the dotted line shows the frequency curve for the Ngapuhi Tribes of the North Auckland Peninsula. This series of 105 should be far more homogeneous than the 421 of the total series, who are derived from 19 different tribes from all parts of New Zealand. The various tribes with different averages and different percentages of long and broad heads have somehow fitted in to make an apparently normal frequency curve. In the Ngapuhi series, however, the different elements come out in the curve. The mean is at 78·6 but the mode or greatest frequency practically coincides at 79. Secondary modes appear at 75 and 81, showing the influence exerted by long-headed and broad-headed elements in the series.
The distribution of long heads and broad heads amongst the North Island tribes varies considerably. My results do not agree with those of Dr. Scott,2 who himself recognised that his data was too scanty to establish the various tribal differences that he believed to exist. Though I have sufficient data for four tribes, it is advisable to deal separately with tribal differences when the other main tribes have been worked up to series of at least 50 subjects. The following table, however, will give some indication of the distribution of head form. The cephalic index is reduced to cranial index and the percentages indicate the percentage within the tribe itself and not in relation to the whole series.
TABLE XXXV.—TRIBAL DISTRIBUTION OF HEAD FORM (Converted).
In spite of the fact that the nose of a new born infant was massaged, the Nasal index still remains a useful criteria in disentangling racial types. It is a question as to how far spasmodic - 192 rubbing without a fixed mechanical appliance would influence the natural growth, development and ultimate shape of the nasal organ. It is generally held that massage by itself has no effect on the ultimate shape of the head. Best3 describes how the nose was pressed flat in a certain district. I myself was told the same thing of the same district, when my Maori informant quoted it as an exception to the rule amongst his own tribe of narrowing the nose in the manner I have already described. Although nose flattening has been recorded for various parts of Polynesia, it appears that in New Zealand, massaging the nose to raise the bridge (theoretically) and narrow the nostrils was the usual thing and flattening was the exception. Whatever effect either method had on the fleshy part of the nose—and I believe that that effect was nil except for the aesthetic pleasure that it gave the mother—it could have no effect whatever on the shape of the nasal aperture of the skull and the situation of the naso-frontal suture. Scott's2 percentages of 43·9 per cent. leptorhine, 45·5 per cent. mesorhine and 10·6 per cent. platyrhine for Maori crania can therefore have no objections raised against them by believers in the maximum effects of massage. My series gives the same percentage of 10·6 for broad noses, so that evidently massage was unable to convert mesorhines into platyrhines. I have lost, however, in the narrow noses for I can only muster 17·9 per cent. for my series as against Scott's 43·9 per cent. If the correllation between the nasal indices for the cranium and the head is correct, this difference will probably be regarded by some ethnologists as proof of the effect of massage in flattening the living nose. It is curious that massage should have changed 26 per cent. of leptorhines into mesorhines and not have converted a single mesorhine into a platyrhine. In any case the two sets are not exactly comparable as Scott's crania included both sexes. I would again stress the point that unless a mechanical appliance is kept constantly in position for a considerable time, massage either to flatten or narrow is useful only to indicate the type of nose desired. Need it always be that of the conqueror or the latest comer? Why should not a conquered people yearn after their own ancestral type of nose that is being altered by miscegenation? A large proportion of Jews7 have maintained their ideal type of nose even when alterations in head form have revealed considerable intermarriage with the surrounding gentiles. May not the different ideals of two distinct ethnic waves, cropping up as one or other racial factor predominates in the tribe, account in New Zealand for the conflicting methods of attempting to flatten or narrow the nostrils.
CORRELLATION.—Correllation of two or more physical criteria has shed considerable light on the existence of different types in a population. Professor Rowland Dixon1 correllating three physical - 193 criteria created eight fundamental ‘types’ for which he worked out the percentages in the cranial data available from various parts of the world. Two of these criteria are the cranial and nasal indices. I have correllated these two indices for this series. His third criteria, the vertical or length-height index does not affect his New Zealand results to any extent. I have therefore omitted it.
TABLE XXXVI.—CORRELLATION BETWEEN CRANIAL AND NASAL INDICES (Converted).
From the above table, it will be seen that by taking the usual three divisions of each of the criteria, nine possible combinations are arrived at. Following the method adopted by Professor Dixon, all combinations which include a medial factor, mesaticephalic or mesorhine are treated as mixtures or blends. These occupy the middle horizontal and vertical columns in the table. Eliminating them for the time being, we are left with four fundamental “types,” one at each corner of the table. Of these four types, two are long-headed and two broad-headed. In each of these pairs, one has a narrow nose and one has a broad nose. The long-headed, narrow-nosed “type” (dolichocephalic-leptorhine) shows the greatest number, with 27. The long-headed, broad-nosed “type” (dolichocephalic-platyrhine) however, is very little behind it, with 25. Both these long-headed types as we would expect from the distribution of the cranial index, are greatly in excess of the broad-headed types. The third type in order of frequency is the broad-headed, narrow-nosed type (brachycephalic-leptorhine) with 10. The fourth corner allocated to the broad-headed, broad-nosed type (brachycephalic-platyrhine) shows a blank.
The four corners of our table, however, account for only 62 of our entire series, whilst 359 have been merged into “blends.” Professor Dixon dealt with these “blends” by resolving them back into “types.” The first vertical column consists entirely of lepto-rhines. Of these, 38 have heads of medium cranial index, which are considered as being a “blend” of long and broad-headed “types.” The 38 are therefore divided between these two narrow-nosed types. Similarly in the third vertical column consisting entirely of platy-rhines, the 20 with medium heads are divided between the two types - 194 above and below them. In the upper horizontal column, consisting entirely of long heads, the 113 with medium noses are divided between the long-headed, narrow-nosed, and the long-headed, broad-nosed types. So also in the lowest column, the 29 broad heads with medium noses are divided between the two broad-headed types. The 159 with medium heads and medium noses could belong to both types of head, and both types of nose, therefore to all four types. Where a type such as the brachycephalic-platyrhine is apparently absent as in this case or has little influence, it is left out of the distribution. Whilst the procedure adopted by Dixon of giving an equal division of blends to the appropriate “types” results in an approximate indication of the percentage of types, it seems capable of improvement in a small series such as this. Where the long-headed factors predominate so much over the broad-headed, it seemed to me that they should have a greater percentage of the medium-headed lepto-rhines and platyrhines than the broad-headed types. Thus the dolichocephalic-platyrhines which number 25 should have more of the 20 mesaticephalic-platyrhines than the brachycephalic-platyrhines which number 0. At the same time the latter, though apparently non-existent as a definite type, should not be entirely disregarded. There are several brachycephalies with a nasal index very little below 85, and there are also platyrhines which just fail to reach the brachycephalic index. I therefore distributed the “blends” amongst the “types” by setting up a dividing index midway in the mesaticephalic and mesorhine groups. For mesaticephalic blends, all up to 79·4 were treated as dolichocephalic, and all above that as brachycephalic. For the mesorhine blends, those up to 77·5 were treated as leptorhine, and those above as platyrhine. The 359 “blends” were distributed amongst the four types according to this method and the results are tabulated below:—
TABLE XXXVII.—PERCENTAGE OF FUNDAMENTAL TYPES (Converted).
The two long-headed types form 73·1 per cent. of the entire series. This preponderance is to be expected when the average cephalic index for the series is 77·7 or reduced to the cranial index with which we - 195 are working, 75·7. Of the two types, however, which were almost equal in the first distribution, the dolichocephalic-leptorhine receives over twice as many accessions from the “blends,” namely 175 to 81. That the majority of long-headed mesorhines should go to the long-headed, narrow-nosed type is again to be expected, when it is noted that the average nasal index of the series is 75·9 and thus nearer to the leptorhine group than the platyrhine. It is unexpected, however, that the majority should be so marked. The long-headed, narrow-nosed type with 48 per cent. is thus easily the most predominant type in the Maori population of the North Island. The long-headed, broad-nosed type maintains its position in second place, but where on the first count it was on almost equal terms for first place, it has now fallen well back with 25·1 per cent. Of the broad-headed types, the narrow-nosed type received an accession of 63, giving it a percentage of 17·3 and thus making it a definite factor not so very far behind the dolichocephalic-platyrhides. The broad-headed, broad-nosed type, which at first had apparently no representatives, received 40 from the “blends,” and is present to the extent of 9·5 per cent.
Reverting back to the first count of types in Table XXXVI., the table below shows the averages for four additional criteria for the three main types, without the addition of the blends.
TABLE XXXVIII.—AVERAGES OF ADDITIONAL CRITERIA.
The average height in the three types is practically the same, though one would have expected to see some more marked difference. In the sitting height index, the dolichocephalic-leptorhines are above the average for the series, which needs explanation if it is supported by further data. In the cephalo-facial index, the brachycephalic-leptorhines show the lowest average, which is to be expected in a broad-headed type. The two long-headed types are both above the average for the entire series and of the two, the broad-nosed by its slightly higher index shows a tendency towards a narrower skull in proportion to its face width than does the narrow-nosed type. In the facial index, there is a very significant difference between the two long-headed types. The narrow-nosed one is 3·4 above the average for the entire series whilst the broad-nosed one is 3·3 below it, making an actual difference between them of 6·7. Thus the long-headed, - 196 broad-nosed type has a distinctly shorter face in proportion to its breadth than the long-headed, narrow-nosed type.
These types are purely laboratory types deduced from a consideration of mathematical figures. Professor Dixon4 found all four types in the cranial material for Polynesia. In combination with the vertical or altitudinal index, he considered them as arbitrary units to which he applied names. The long-headed, narrow-nosed type he found predominant in New Zealand. This type he associates with the Causasian factor that various writers have stated to be present amongst the Polynesians. In his more recent work1 he distinguishes the one with a high head as Caspian, and the one with a low head as Mediterranean, but remarks that they are found associated together. It is the former that he finds predominant in the North Island of New Zealand. The long-headed, broad-nosed type, when associated with a high head, he terms Proto-Negroid, and with a low head, Proto-Australoid. A mixture of the two, he associates with Melanesians and finds occupying second place in the North Island. The broad-headed, narrow-nosed type, he termed Alpine, and placed in the third place in the North Island and in the first place in the South Island and the Chatham Islands. Our data, reduced to the same terms of cranial index and with the “blends” distributed amongst the “types,” corroborates Dixon's views for the North Island as a whole, but do not agree with his division of the North Island into two areas, the North Auckland Peninsula and the rest of the North Island. (See Table XXXV.) The broad-headed, broad-nosed type with high heads which he found predominant in the northern Islands of the Sandwich Group, he found existing as a mere trace in other parts. This type he first termed as Negritto,4 but from the consideration of further data, he changed the term to Palae-Alpine.1 In our data, though there were no pure examples of this type, the distribution of the “blends” shows that as an arbitrary unit it exists in the Maori population to a greater extent than would have been expected.
Dr. Sullivan,5 from the more extensive and detailed Polynesian living material rendered available by the Bayard Dominick Expedition, regards the Polynesian population as consisting of at least two basic elements. One with tall stature, moderately long heads, relatively high narrow faces and relatively high narrow noses, he considers as the Polynesian proper. This he considers probably agrees with Dixon's Caucasian type (subsequently Caspian). It would also agree with the dolicho-leptorhine element, with a relatively high face which in Table XXXVII. formed the predominant element in our Maori series. The other element with shorter stature, shorter heads, low broad faces, and low broad noses, he terms Indonesian. Though - 197 to many, the term Indonesian has usually been associated with long-headed races living in Indonesia amongst short-headed races, Sullivan's enumeration of the physical characteristics leaves no doubt as to the type he means. He says that this is probably the element that Dixon terms Negrito (and subsequently Palae-Alpine). It probably coincides with our brachycephalic-platyrhine element, which we have found to be present in our Maori material to a small extent. If the Maori material is analysed without reducing the cephalic index for comparison with Dixon's cranial results, this second element is found to be present to an even greater extent.
TABLE XXXIX.—CORRELLATION OF CEPHALIC AND NASAL INDICES. (Unconverted.)
The difference of two units, as already pointed out, considerably alters the proportion of long-headed and broad-headed elements. Amongst the platyrhines, the broad-headed element, previously apparently non-existent in a “pure” type, now exceeds the long-headed element. If the “blends” are further distributed by dividing the mesaticephalic at 77·4 and the mesorhines as before as 77·5, the difference between broad heads and long heads is still more marked.
TABLE XL.—FULL DISTRIBUTION OF ELEMENTS. (Unconverted.)
This brings the broad-headed, broad-nosed type into consideration as a definite, if small element in the Maori population. My data was collected on a troop-ship before I knew of the organised work being done by members of the Bayard Dominick Expedition. Many of the general characteristics that Dr. Sullivan has used to distinguish his Indonesian element were not recorded by me. This can be rectified by subsequent observation. I have, however, no doubt in my mind - 198 that a certain proportion of the Maori brachycephalic-platyrhines will correspond to Sullivan's Indonesian element and that though it does not appear to nearly the same extent as amongst the Polynesian material with which he has been dealing, yet it does exist amongst the Maoris.
Sullivan5 recognises a third element characterised by extremely short heads, narrow faces and narrow noses. This he was inclined to regard as a Polynesian type with artificially deformed heads and as perhaps corresponding to Dixon's Malay4 and subsequently Alpine1 type. It is now evidently regarded by Sullivan6 as perhaps forming a distinct element which probably contributed some of the Caucasoid traits to the Polynesians. Though Table XL. brings the brachycephalic-leptorhine element into the predominant position, a large proportion of it does not have a very short head. Taking the cephalic indices as against the converted cranial indices, a broad-headed, narrow-nosed element undoubtedly exists in the Maori data to a very important degree.
The long-headed, broad-nosed element exists in the above table to the extent of 18·9 per cent. It has fallen from second place in Table XXXVII. to third. Sullivan recognises this element in the population of Polynesia, and like Dixon calls it Melanesian.
Much vexation of spirit has been caused by changing about from skulls to heads. This could not be obviated in this article as we had to compare our results with that of two authorities working with different material. This unnecessary amount of extra work and alternate summings up will continue to exist as long as the figures dividing the cephalic and cranial indices into their three classes, remain the same. In the data supplied in this paper, the broad-headed elements predominate over the long-headed, but when converted for comparison with cranial material, the position is reversed.
In conclusion, as an interim report, we may say that four elements can be recognised amongst the Maoris. Of these four, two with narrow noses easily predominate over two with broad noses. In the living, the broad-headed, narrow-nosed element is, contrary to expectation, the most important. The second or long-headed, narrow-nosed element is so little behind the first in frequency, that the difference of two units in converting the cephalic into cranial indices, reverses their position.
The two broad-nosed elements, with long and short heads, though not nearly so prevalent in the whole population occur, nevertheless, as distinct elements in certain areas.
Marked tribal differences exist owing to the varying proportion of these elements. Their association with definite types is postponed until further details as regards general characteristics can be added. - 199 Their association with the waves of migration of traditional history, though tempting, is also deferred until gaps in intertribal material have been supplied.