The human remains from Keiss submitted to me by Mr. Laing consist of the greater part of one skeleton with many portions of others; and comprise nine skulls and fragments of skulls, with four more or less complete pelves belonging to certain of the skulls, besides bones of the limbs and of other regions of the body.
What I have to say respecting them will be arranged under the following heads:
I. A description of the human remains from Keiss.
II. A comparison of these remains with one another and a discussion of their ethnological characters.
In describing the crania I shall adopt, in principle, and mainly, in detail, the convenient nomenclature which has been gradually developed by Retzius, Broca, Welcker, and Thurnam, and which is employed by the latter writer in his valuable essay "On the Principal Forms of Ancient British and Gaulish Skulls." The term cephalic index, in this nomenclature, indicates the ratio of the extreme transverse to the extreme longitudinal diameter of a skull, the latter measurement being taken as unity. Skulls with a cephalic index of 0.8, or more, are Brachycephalic,1 or "round." Those  with the cephalic index less than 0.8 are Dolichocephalic. It is in this sense that the terms dolichocephalic and brachyocephalic were used by their inventor Retzius, so far as he defined them at all:
"Vous me demandez les caractères distinctifs entre la forme brachycéphale et dolichocéphale! Je ne veux pas encore determiner quelques mesures fixes pour les distinguer; mais, à l'ordinaire, le diamètre longitudinal des dolichcéphales surpasse la largeur d'environ 1/4, tandis que chez les brachycéphales cette difference varie entre 1/5 - 1/8." ("Ethnologische Schriften," p. 118).
As Retzius considers the Swedish skulls with an index of .773 (l.c. page 4) to be eminently dolichocephalic, he must mean by the above expression, that, in a dolichocephalic skull, the long diameter exceeds the short by at least a fourth of that short diameter. In other words that the length is, at least, as 1000: 800. If he had meant that the long diameter exceeds the short only by one-fourth of the long diameter, no skull would have been dolichocephalic unless the long diameter had to the short the proportion of 1000: 750, a definition which would exclude his Swedish skulls from the dolichocephalic group. For this reason I must dissent from the interpretation which the editor of Retzius' Ethnologische Schriften puts upon his author's words (l.c. page 121, note). But apart from any question as to what may have been Retzius' meaning, the terms brachycephalic and dolichocephalic have been so extensively used, in this country and elsewhere, to denote skulls with the cephalic index above and below 0.8 respectively, that great inconvenience would result from attaching any other signification to them.
While objecting, however, to the use of "dolichocephalic" in any new sense, I quite agree with Broca, Welcker and Thurnam in thinking it expedient to sub-divide the dolichocephalic division; and, in doing this, I shall adopt the very convenient grouping suggested by Dr. Thurnam, who arranges the cephalic indices below .80 into four groups:
|(.79 .78 .77);||(.76 .75 .74);||(.73 .72 .71);||(.70 and below).|
To No. I. Dr. Thurnam gives the name sub-brachycephalic used by M. Broca, and which I shall adopt, though not without some reluctances on account of its etymological hybridity.
No. II. is the orthocephalic division.
No. III. is called subdolichocephalic, and No. IV. dolichocephalic, by Dr. Thurnam; but I object to both names for the reasons given above, and I substitute for No. II. the title of mecocephalic, and for No. III. mecistocephalic. Finally, I propose to sub-divide the Brachycephali into Eurycephali, with the cephalic index .80 to .84, and Brachistocephali, with the cephalic index .85 and above. The following table will make this terminology plain, and exhibit the relations to that already in use:
|Ophalic index,||at or above .80||=||I Brachyccephali,||round skulls.|
|"||at or above .85||=||Brachistocephali,|
|"||below .85 and at or above||.80||=||Eurycephalic.|
|Ophalic index,||below .80||=||II. Dolichocephali,||long skulls.|
|"||below .80 and at or above||.77||=||a. Sub-brachycephali,||} oval skulls.|
|"||below .77||"||.74||=||b. Orthocephali,|
|"||below .74||"||.71||=||c. Mecocephali,|
|"||below .71||=||d. Mecistocephali,||oblong skulls.|
The Skull No. 7
In the lateral, front, and back views of this and all other skulls figured, Mr. Busk's method of placing the skull has been adopted. A plane traversing the auditory meatuses and the junction of the coronal and sagittal sutures is vertical. In the "norma verticalis" the centre of such a plane is traversed by the line of sight. All the drawings are reduced to one-third of the natural size of the skulls.
I. A Description of the Human Remains from Keiss.
I now proceed to describe the six skulls from Keiss and the other bones associated with them, indicating each set by the numbers affixed to them by Mr. Laing.
1. The Bones marked No. 7
The skull (figs. 1 to 5), though slightly damaged, is in a  very fair state of preservation, and is a remarkably well formed and spacious cranium. The cephalic index is 0.78, and it therefore belongs to the sub-brachycephalic division.
The nasal depression and the supraciliary prominences are moderate; the forehead is well arched, though not high, and rises almost vertically from the brows. The superior curved line is well developed, and the occipital spine is strong; but the broad and convex occiput projects beyond it, when the glabello-occipital line is made horizontal.
The coronal suture is traceable throughout; the sagittal and the middle part of the lambdoidal are almost completely obliterated. A flat, broad depression runs along the middle line of the skull, in the region of the posterior half of the sagittal suture (fig. 2).
The upper contour of the squamosal is nearly straight; the mastoid processes large and prominent; the nasal bones well developed and projecting. The upper jaw is massive, with a sharp and strongly marked spina nasalas anterior. But there is no prognathism and the outer surface of the alveolar region is almost perpendicular. The lower jaw has very massive horizontal and perpendicular branches, and the mental prominence is so well marked that, when the zygoma is horizontal, the chin projects as far as the level of the extremities of the nasal bones.
In the upper view (or norma verticalis) of the skull (fig. 2) its broad and rounded figure is well displayed, and it will be observed that no part of either the jaws or the zygomata is visible. The front view (norma frontalis) (fig. 3) exhibits the massiveness of the jaws, which yet in no vise predominate over the brain case. The upper jaw, in fact, is not so round and full in front as might be anticipated from the general mass of the facial region. The arch of the palate is consequently (fig. 5) particularly acute, and the fore part of the face looks, as it were, pinched. In accordance  with this, the nasal bones are prominent, and meet at an acute angle, so as to form a dorsal ridge in the upper part of their extent. The back view (norma occipitalis) (fig. 4) shows the rounding off of the parietal and sagittal regions, so that the contour is more arched than pentagonal, and, furthermore, reveals a slight want of symmetry in the two sides. The palate (fig. 5) is remarkably deep and rather narrow, measuring 1.95 in. from the incisor alveolar margin to the level of the posterior boundary on each side; 1.5 in. opposite the two last molar teeth where it is broadest; and having a depth of 0.67 in. from the roof to the alveolar margin, where it is deepest, which is at a point close to the posterior edge.
In these and all the other figures of pelvis, the front and side views are taken from the pelvis in such a position that the body of the ischium is vertical. The figures are one-fourth of the natural size.
The occipital foramen is directed downwards and forwards. The mastoid processes are large, the supra-auditory ridges strong, and the styloid processes are strong and well ossified. The left upper canine is thrust inwards out of its place, while in the lower jaw the teeth are regular. In both jaws, the enamel of the incisors and of the molars is worn away so as to expose the dentine, but there is no symptom of decay in any of the teeth.
The sacrum and three fragments of the left innominate bone belonging to the same skeleton as the skull No. 7 were obtained by Mr. Laing. When they were carefully put together, and the right innominate bone was restored after the model of the left, the result was the pelvis represented in figs. 6-8, which are reduced to one-fourth of the natural size. The transverse diameter of the brim of the pelvis is, as nearly as I can judge from measurement of the specimens 4.6 in.the figure making this diameter slightly too small. The oblique diameter is equal to the transverse. The promontory of the sacrum is broken away, whence the precise ascertainment of the antero-posterior diameter becomes impossible, but it cannot have exceeded 3.8 in. The sacrum is markedly concave forwards, the pubic angles acute, the descending rami  of the pubis are straight and not arcuated, and the true pelvis is deep. In fact, this is a well marked male pelvis differing in no important respect from the ordinary pelvis of the European male, though it is rather smaller than the average. From the measurements of the sacrum and those of the acetabulum, I conceive that the stature of No. 7 must have been below, rather than above, the middle height, perhaps five feet four or five inches.
2. The Bones marked No. 8.
The skull is orthocephalic, the cephalic index being 0.76. The calvaria is remarkable for the projection of the supra-ciliary ridges, caused by the great development of the frontal sinuses, the prominences formed by which are separated by but a very slight median depression. The forehead is low, narrow and receding, and the median contour of the skill slopes from it, gradually upwards, towards the vertex. The occipital spine is very strong, curved, and prominent, and the cerebellar region below it, strongly convex downwards. The supra-auditory ridge and the mastoid processes are very strong, and the styloid processes seem to have been no less so.
The norma verticalis (fig. 10) shows that the skull is narrower in the frontal region and widens out towards the parietal tuberosities. The sagittal and lambdoidal sutures are open throughout; the coronal is open above, but is obliterated for about an inch and a half at each of its outer ends. There is a large Wormian bone in the right crus of the lambdoid suture. The norma occipitalis (fig. 12) shows that the transverse contour of the skull inclines to be pentagonal and wall-sided.
The facial bones of this skull were entirely detached, and, like the base of the skull, much broken. However, of the lower jaw rather more than the left half was entire; and  of the face the left jugal, left maxilla, and part of the right maxilla were preserved. Consequently, by articulating the mandible, making its median line coincide with that of the skull, and then adjusting the teeth of the fragmentary upper jaw to those of the lower, and the jugal to the external orbital process of the frontal, it was possible to restore the left half of the face, as in fig. 9. The prognathism of the skull thus restored is remarkable, especially since the front walls of the alveoli, in the premaxillary region and in the front part of the lower jaw, are broken away; the front tooth shown in the figure, being the canine in the upper jaw, and the outer incisor in the lower.
The axis of the foramen magnumthat is, a line drawn perpendicularly to its plane and through its centreof this skull is directed, not downwards and forwards, as in human skulls in general, but downwards and backwards.
The sacrum, though it consists of only the ordinary five vertebræ is as long as it is broad, and has a well marked anterior concavity. The greater sacro-sciatic notch is not more open than usual; the true pelvis is deep, and all the ridges and tuberosities of the bones are well developed. The axes of the acetabula are directed well forwards as well as outwards (figs. 14, 15, 16).
The vertebral fragments consist of the neural arches and attached posterior segments of the bodies of the second and third lumbar vertebræ The fragments are naturally adjusted, and the distance from the upper face of the body of the one to the lower face of the body of the next, under these circumstances, is 2.6 in., which gives a height of 1.3 in. for each vertebra. The spinous processes, transverse processes, metapophysial and anapophysial tubercles are all remarkably strong and well developed.
The other bones belonging to No. 8 are the greater part of each innominate bone, the right and left femora  and tibiæ the two humeri, minus their proximal ends and some fragments of the fibulæ and lumbar vertebræ. On carefully joining the sacrum with the fragmentary innominate bones, the result was the pelvis represented it figs. 14-16. As the pubis and the amending ramus of the ischium of each side are wanting, all the measurements of the brim cannot be ascertained. Its transverse diameter, however, is 4.7 in., and from the direction of what is preserved of the lateral boundaries (fig. 16), I do not believe the conjugate diameter exceeded 4.0 in.
There is nothing worthy of special notice about the other bones, except that they are strong, and have all their processes and ridges well developed. The femur is 18.7 in. long, which, by computation, would give the stature at 5 ft. 8 in. But the tibiæ are only 14.7 in. long, or nearly an inch shorter than they should be, in proportion to the length of the thigh bone. This deficit may have been made up by an unusual length of the strong spine, but it is quite safe to assume sixty-eight inches as the maximum height which this individual can have possessed.
 For anything we know, or have other than philological reasons for believing to the contrary, fair brachycephali ("Belgæ," Fins, etc.); fair dolichocephali (Scandinavians); dark but pale-complexioned dolichocephali (Iberians, etc.); and dark and yellow skinned brachycephali (Lapps), may have inhabited some part or other of the area they now respectively occupy in Europe for long ages before the dawn of history.
Hence, it is worth while to reflect, that the current notions respecting the migrations of races from east to west may be myths developed out of the facts of philology; and that successive waves of language may have spread over Europe by washing over, instead of being carried by, its populations
That this is what has happened and is happening in our own islands appears to me sufficiently probable. Our population contains three distinct ethnological elements: I. Xanthochroi brachycephali; II. Xanthochroi dolichocephali, and III. Melanochroi. In Cæsar's time, and for an indefinitely long preceding period, Gaul contained the first and third of these elements, and the shores of the Baltic presented the  second. In other words, the ethnological elements of the Hiberno-British Islands are identical with those of the nearest adjacent parts of the continent of Europe, at the earliest period when a good observer noted the characters of their population.
Dr. Thurnam has adduced many good reasons for believing that the "Belgic" element intruded upon a pre-existing dolichocephalic "Iberian" population; but I think it probable that this element hardly reached Ireland at all, and extended but little into Scotland. However, if this were the case, and no other elements entered into the population, the tall, fair, red-haired, and blue-eyed dolichocephali, who are, and appear always to have been, so numerous among the Irish and Scotch, could not be accounted for.
But their existence becomes intelligible, at once, if we suppose that long before the well-known Norse and Danish invasions, a stream of Scandinavians had set in to Scotland and Ireland, and formed a large part of our primitive population. And there can be no difficulty in admitting this hypothesis when we recollect that the Orkneys and the Hebrides have been, in comparatively late historical times, Norwegian possessions.
Admitting that in the prehistoric epoch, central Europe was peopled by short-headed Xanthochroi; northern (Baltic) Europe by long-headed Xanthochroi; and Western Europe by dolichocephalic Melanochroi, the present and past states of the population of the same area become intelligible enough.
In ancient times when, to use Dr. Dasent's words, "Scandinavia was the great slave market of Europe," the introduction of fair brachycephali into the Baltic area may as readily be understood (without having recourse to any special "Finnic" hypothesis) as the elimination of this element, and the return of the Scandinavians to the long-headed type, in modern times, when the brachycephalic infusion ceased.
In another fashion, the fair and broad-headed "Belgæ" intruded into the British area; but, meeting with a large dolichocephalic population, which at subsequent times was vastly reinforced by Anglo-Saxon, Norse and Danish invasions, this type has been almost washed out of the British populations which is, in the main, composed of fair dolichocephali and dark dolichocephali.
The reverse process has obtained in central Europe. When the great Teutonic stocks swarmed into the Roman Empire, as the Gauls, with less success, had attacked the Republic, they spread the type of the dolichocephalic Xanthochroi far beyond its primitive bounds. But, however they might seem to be conquerors, the Franks and Alemanni, who settled in central Europe, were ethnologically defeated. On their right flank were more numerous "Belgæ" and people of like stock; on the left flank innumerable Slavonians. Under these circumstances, while complexions might remain unchanged, dolichocephaly had no chance against brachycephaly, and accordingly the latter has eliminated the former.
But language has, in no respect, followed these physical changes. The fair dolichocephali and fair brachycephali of Germany, Scandinavia, and England speak Teutonic dialects; while those of France have a substantially Latin speech; and the majority of those of Scotland and within historic times, all those of Ireland, spoke Celtic tongues. As to the Melanochroi, some speak Celtic, some Latin, some Teutonic dialects; while others, like the Basques (so far as they come under this category), have a language of their own.
1 "Brachycephalic" is of course in strictness applicable to the people to whom the skulls belong; but custom has sanctioned the looser use of the word.
1 "Brachycephalic" is of course in strictness applicable to the people to whom the skulls belong; but custom has sanctioned the looser use of the word.