MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



215 



In making: the- transfers to paper superimpose one sketch on the other in such a way that the 

 axial lines shall cross or touch, and the parallel lines shall exactly coincide. Apply the protractor 

 and read off the angle of torsion.* 



In every case where we have applied this method we have, as a matter of record and identi 

 fication, drawn the outlines of the extremities, an easy task ; but it would have been sufficient to 

 draw only the axial and parallel lines. 



Figure 45 shows the character of the tracing. The outline of the head is broadened in order 

 to distinguish it more plainly from that of the opposite extremity. 



The general results of our measurements are shown in Tables LV to LVIII, inclusive, and in 

 diagram shown below. Tables LIX and LX give the angles of other humeri in our Museum. A 

 number of tables prepared by Broca and Manouvrier have been 

 consulted which, though the measurements were taken by a dif 

 ferent process, will, we believe, admit of comparison with our 

 results. From all these sources the following facts are collated: 



A statement of Broca s,t based upon abundant data, is that the 

 average torsion is greater in females than in males, and his Table 

 D shows that not only in the general average, but that in the 

 average for each side the female exceeds the male, there being 

 but one insignificant exception in the series of Californians. In 

 this respect the Saladoans seem to be at variance with the rest 

 of the human race. In Table LVIII it will be observed that hu 

 meral torsion in the males is greater on both sides, and therefore 

 greater in the total average, than it is in the females. 



Another conclusion of Broca sJ is that in nearly all the series 

 (studied by him) the left humerus is, on an average, inore twisted 

 than the right; such, too, is the evidence of our general collection 

 (Table LX) even with regard to American races. In the Salado 

 skeletons, on the contrary, the average is almost the same on both 

 sides, that of the right being slightly in excess of that of the 

 left. Among the humeri in pairs, also, there is a slight excess on 

 the right side. The variation, too, is greater on the right than 

 on the left side in this series, the former showing both higher and 

 lower angles than the latter. 



In 75.8 per cent of Broca s series the maximum of torsion is 

 on the left side. Here again the Salado series ranges itself 

 with the small minority. Not only the maximum but the highest 

 three angles are found on the right side. It belongs to the majority, however, with regard to the 

 minimum, which is on the right side as in 72.4 per cent of Broca s series. 



In comparing the, liumeri of this series (Table LV) with those of our general series (Table LTX) 

 we discover that three angles of the former (177, 174, 174, all dexter) are higher than the maxi 

 mum of any other race except the French, and that they are higher than several of the French 

 angles. If we study this series in connection with IJroca s Table C, || in which is given a list of 2!) 

 series, comprising the most varied races in the world, the maximum angle of the Saladoans would 

 still seem to have the same relative importance standing next to the. French. 



The average torsion of the left humerus (159 30 ), the average, of the right humerus (159 45 ), 

 and the average of all the humeri (159 30 + ) are higher than the corresponding averages in any 

 series (representing more than one individual) of our collection except the French and Lapps. 



Fm. 45. Tracing showing torsion of hu- 

 merlin. 



&quot;Journal of Anatomy uml Physiology, vol. xxr, ]&amp;gt;. 536. 



t I. a lorsicm &amp;lt;lc I limnerus. Revue d Anthropologie, 2&quot; Sfirie, T. 10, 1 aris, 1881; pp. 577 et arq. 



tZoc. cit., p. 383. 



$ /. . ,-il., p. 584. 

 || J.ov. ci/., ]&amp;gt;. .&quot;&amp;gt;s:(. 



