ARCHAOLOGICAL AND ETHNOLOGICAL RESEARCHES IN CRETE. 261 
I have supplied the data necessary for extending tests of this kind 
in the next table (XXI.) where the standard deviation, the coefficient 
2 
of variation, and the numerical value of the ratio 2 are recorded for 
N 
each of the measurements made on the Reformatory youths. 
Table XXII.—The correlation of the length with the breadth of 
the head is extremely low. The comparable data for the height and 
length, as for the breadth and height, are in an ascending scale, so 
that the sequence runs as follows :— 
1. Length and breadth (least closely related). 
2. Length and height. 
3. Breadth and height (most closely related). 
But the correlation is of a low value throughout. The value of the 
coefficient for the cephalic index and the length is more interesting. 
As the table shows, this is of high value, but with the negative sign 
(—0°63). That this should be the case could be inferred from what | 
have stated in regard to the correlation of the length and breadth. It 
is, however, of interest to compare this value with those derived from 
three sets of skulls presenting marked dolichocephalic proportions. 
Thus we have 
Index 
Greek heads SS ee eee ee cet ey se OO (S05) 
Sardinian crania ee A MET EPO 0543? CUS) 
Naqada Egyptians . . . . . . .—0-551 (72°99) 
Medieval English . 2 i ‘ : ; . —0-547 (74-34) 
Correlation here seems to be comparatively independent of the absolute 
value of the cephalic index. 
C. Distribution of the Cephalic Index and the Nasal Index, as shown 
on the Maps. 
The regional distribution of the young Greeks is now to be con- 
sidered. Their great variability in respect of the cephalic index has 
been noticed already. Hitherto I have dealt with them as a single 
group, but it is necessary to add that they come from a wide area, and 
that, of the ninety-nine, fourteen are from islands close to the main- 
land, while the remote islands supply four individuals. They are 
classified as follows :— 
1 1SR 4 - 3 5 j . Mainland 81 
Tonian Islands 8 
II. Islands close to mainland : | Eubea . 4 
Spetzai . MES MH TER Re 
III. Remote islands . . . Amorgos, Mitylene, Naxos, 
Paros, one each 3 ae ee 
99 
With regard to the cephalic index, I have drawn up the comparisons 
set forth in Table XXIII. 
