350 



Miscellanea 



We have the following data for males : 



Now it is clear tliat thesc resvilts fall very wide of the mark — giving with one exception 

 values luostly far too small ; or we coiiclude that : Xo satisfactüry interracial formulae hy which 

 either zygomatic or cephalic breadtlis could be deduced from frontal breadth can be obtained even 

 from data as varied as tliat provided bi/ the tribes and castes of Bengal. 



While an interracial formula based on that data gives fairly good results for a special trihc it 

 cannot be applied to such diverse matorial as the Naqada, l'avarian, and Aino series. In fact 

 our Bengal material ha.s thronghout a remarkably small cephalic breadth as compared with the 

 frontal and zygomatic breadths, and we get formulae practically of no service at all for such 

 crania as the Germau or Aino. We have taken too limited a group of human races. To 

 obtivin interracial formulae we must clearly proceed «n a much wider basis, and such an investi- 

 gation we hope to uudertake at a later date. 



In the next place we turn to an iiidividual race and write down the cquations by which 

 the probable cephalic [p^) and bizygomatic {p^) breadths might be predicted for an individual 

 head of this race. 



For the Statistical constants of the Orann tribe given above we have : 



j»2 = 63-21 + -65954/; 

 jo<,= 55-45 + -82245/'. 



For a second race we take 52 $ Naqada skulls for whii-h the three breadths C, F, and ^are 

 known. We lind : 



i/,.= 133-68, J/,.= 90-52, J/2 = 125-39, 



o-c = 4-3596, o-, = 4-4364, o-^ = 5491 4, 



r„= -5695, »•k.= '0440, r„= -1107. 



The corresix>nding formulae for reconstructing the bizygomatic and cephalic breadths from 



the iiiinimum fi-ontal breadth aro : 



jt)2= 61-58 + -70497/', 

 jOc=123-84 + -10875/'. 



The marked feature of the Naqada results is the great drop in the correlations between the 

 cephalic and zygomatic and tetween the cephalic and the frontal l)readths whcn we pa-ss from 



