There are three methods of considering the trunk as an object of measurement. The first of these is to take the measurements of the spinal column from the first dorsal vertebra to the termination. The second, which is strictly anatomical, is to disregard the clavicles and other portions of the shoulder-girdle above as belonging properly to the upper extremity, and to confine the measurements to the thoracic, abdominal, and pelvic cavities. The third, which is certainly the most useful from an artistic point of view, is to include the portion omitted in the second, and to measure the trunk as it appears to exist in the nude figure clothed with its skin and muscles. This is the system adopted by the French Society of Anthropologists, whose directions state that measurements are to be taken from the suprasternal notch to the seat, that is, to that portion of the body which rests upon the ground, or upon a chair, in the sitting posture. Topi-nard's conclusions drawn from measurements made in various ways are as follows: The relation of the trunk, considered as the vertebral column, varies within narrow limits, as Carus, who on this account took it as his standard of comparison, had already pointed out.

At the same time differences do exist; thus, the Esquimaux and the Tasmanians, so far as the measured cases go, have a trunk shorter than the average; the Samoyedes, the Indo-Chinese, the Polynesians, and the South Americans, all yellow races, have one which is longer. The mean of 108 Europeans examined being 33.8, we may say, in order to assist the memory, that the average of humanity is 33.33, or one-third of the stature. The character seems to vary somewhat in different races; but amongst Europeans the female has proportionately a longer trunk. Finally summing up all the evidence which he has been able to obtain from various sources, he concludes that in whatever manner measurements be taken, provided that similar observations be compared with one another, the trunk will be found to be longer in the yellow races, shorter in those of the negro type, and of intermediate length in white races, though exceptional cases are met with in each which are contrary to the rule. The female has a longer trunk, at least amongst European nations; and in individuals of lofty stature the trunk is longer. Turning now to Marshall's directions, it will be found that he agrees with Topinard in making the female trunk longer, and this for reasons which I have already detailed.

His measurements are taken to a point one and a half units below the tuberosity of the ischium, which is the bony point on which the body rests in the sitting posture, and we must, therefore, subtract this figure from the twenty-seven units which he allows for the trunk and neck. We have seen that he allows three of these for the neck, thus four and a half must be deducted in all. If we reduce the figure of twenty-two and a half units thus obtained to the same terms as those employed by Topinard, we find that it comes out as 33.5, which is very nearly the figure given by the French author.

The intrinsic measurements of the trunk have also been dealt with by Marshall, and the following are his principal results. In the transverse direction the measurement from one deltoid; prominence to the other - that is, the extreme breadth of the shoulders in the nude subject - is in the male two heads; that is exactly one-half of the length of the axis, that from one acromion process to the other, or the maximum breadth of the skeletal shoulders, being one unit less. In the female the deltoid measurement is seventeen units, or one unit less than two heads, and thus it is proportionately shorter than in the female. The distance between the nipples is one head in the male, one unit less - or eight units - in the female. The normal waist in both sexes is ten units, being thus one sixty-seventh of the stature more than a head. The width of the brim of the pelvis is eleven units, and the measurement across the trochanters the same in the male, whilst in the female these two figures are twelve and a half and fourteen and a half units respectively.

The following table gives three of the more important antero-posterior measurements in the male and female respectively:

Male.

Female.

Level of nipples

10½

Waist ..........

8

8

units.

Gluteal prominence

9

10½

From these measurements it will be obvious that in the male the transverse and antero-posterior diameters are greater above - that is, in the region of the shoulders - than below, in the region of the hips; whilst in the female, though the superior are also greater, the difference is not so marked as in the male. On this important point Duval has the following remarks, which I think worthy of quotation. Comparing the diameter of the hips with that of the shoulders, he says: 'What strikes us most in this comparison, at the first glance at a series of skeletons, is the great projection which the hips form in the female. In order to express this, various formulae have been proposed. They consist in considering the trunk as a figure more or less regularly oval, of which one extremity corresponds to the shoulders, the other to the hips, and in determining, according to the sex, which diameter exceeds the other. The ancients did not hesitate to express this formula in the following manner: In the male and in the female the trunk represents an ovoid - that is to say, an oval similar to that of a figure of an egg, having a greater and a lesser extremity; but in the male this has its greater extremity superior, while in the female the greater is inferior.