1887] Variation of the Human Shoulder-Blade. 629 
the indices do not necessarily indicate the shape of the bone, and 
that they are worthless to determine the race of any single bone. 
The one hundred and thirteen bones which I have called Cau- 
casian are, like those used by Flower and Garson, rather a heter- 
ogeneous collection. They for the most part belong to the 
Harvard Medical School and to the Boston Society of Natural 
History. More than a few of them came from France. While 
they, no doubt, are in the main Caucasian, it is probable that 
there are some negro bones among them. Indeed, I know that 
two of the scapulæ came from the body of a negro. These are 
remarkable as presenting very low (instead of high) indices. 
Through the kindness of Professor Putnam, who most gener- 
ously put all the stores of the Peabody Museum of Archeology 
at my disposition, I had hoped to be able to present a large col- 
lection of figures from the bones of the mound-builders, and per- 
haps to make observations on many individuals of a race which _ 
was less mixed'than most of those whose bones are easily ob- 
tained to-day. I was, however, disappointed, and from a cause 
that is easy to foresee, namely, the great fragility of the scapula. 
I could have had long bones in abundance, but the shoulder- 
blades were for the most part either in fragments, or so injured 
that the necessary measurements could not be made. I have the 
records of six scapulz from California which are probably Indian, 
and of eighteen from the Kentucky mounds, The mean of the 
Californians is 67.25 for the scapular index and 91.05 for the . 
_ infra-spinous. The mound-builders have a mean scapular index 
of 69.29 and an infra-spinous one of 93.75. 
The following table shows the indices of Caucasian bones ob- 
tained by my predecessors and myself. Livon is the only one 
who has a tolerably large series showing the difference between the 
sexes. Broca, with a much smaller series, had different results: 
TABLE I. 
Broca. a — Dwight. 
s Male. Female. 
Scapular index...... 65.91 65.2 63.0 67.4 | 63.5, 
. nee de: 87.79 89.4 85.4 88.8 85.8 
To understand the significance of figures something more io 
needed than an average, which gives no hint of the range of © 
Paeon, and — I have arranged the ae of ibe in- 
