98 
The Inhabitants of the Eastern Mediterra^iean 
error is so small that for the moment the matter can be left open. Grouping our 
mean values we have three classes : 
That such a grouping is of any real significance seems unlikely when we turn to 
the standard deviations. Here we have a striking contrast between Lycian Greeks 
and all the other Greek countries, suggesting a very big mixture of two races. 
Apart from the figui'o from Meligala in Messenia where the numbers measured are 
too small for any satisfactory review of the standard deviation it will be seen that 
in every case we have a much greater standard deviation than in the comparatively 
homogeneous Corsica and greater than in the cosmopolitan Alexandria. It should 
also be noticed that the smaller series have smaller standaixl deviations suggesting 
that the small areas from which the men were drawn are more homogeneous than 
the larger areas as for instance Crete or Cyprus. The evidence of four areas 
measured in Cyprus does not however bear this statement out, as the difference 
between the Cypriot standard deviations either between the villages themselves, or 
between the villages and the total number of men measured, is seldom of any 
significance. It may be concluded that we are dealing with a very mixed population. 
This fact is strikingly confirmed by two pieces of evidence. First, the local 
differences between different groups of villages in Cyprus. The villages on the 
north coast (2'21 adult males) have a cephalic index of 81"94 and 167 adult males 
from villages round Enkomi one of 83"3S. Secondly, 60 adult males from five 
selected eparchies of Crete have a cephalic index of 76"6, and 72 adult males from 
another five eparchies a cephalic index of 818. The selection of eparchies by 
V. Luschan on the basis of cephalic index naturally increases the difference and 
tends to emphasize the third piece of evidence more strongly than it perhaps 
deserves. Even if we disregard it however there appears a good case for a great 
deal of admixture among the living " Greeks " and at the same time considerable 
local variation and the appearance of similar types at opposite ends of the 
Greek world. 
One further point which tends to suggest mixture and proves the necessity of 
taking large series is the great contrast between Hawes' and v. Luschan 's figures 
for Sphakia and Selinos. The cephalic index for the whole series of Cretans is the 
same in both cases and diff"ers by only '06 per cent., whereas however v. Luschan 
believes that the cephalic index of the Sphakiots (30 measured) and Selinots 
(21 measured) to be 81'6 in both cases. Hawes, on the evidence of 98 and 35 
respectively, has a cephalic index of 85. The very close resemblance of their final 
mean suggests a similar technique, but this immense local divergence shews clearly 
the danger of elaborating racial theories on an insufficient number of cases. It also 
throws very considerable doubt on v. Luschan's cephalic index map of Crete and 
any conclusions that may be drawn from it. 
Under 81 
82 
Over 84 "o 
Crete 
Mani 
Lycia (Greeks and Turks) 
Meligala 
Cyprus 
Leukas 
Albania 
Lycian Gypsies 
