L. H. DuDLF.Y Buxton 
103 
series are full}' worked out or other data are available from elsewhere this is difficult 
to test on the series under examination. For small series the difference between 
V. Luschan's measurements on 49 men and Hawes' Sphakiots and Selinots (98 and 
33 men respectively) is not great. The latter are woi'ked out on absolute meastire- 
ments and we give below v. Luschan's absolute measui'ements index. Compai-ing the 
absolute measurements Hawes' and v. Luschan's means of upper facial height agree 
to within a millimetre and the divergencies occur in the bizygomatic width, but 
even here the difference — under 3 mm. — is not too great to be the I'esult of chance. 
The figures are : 
Place 
Authority 
Bizygomatic 
Width 
Upper Facial 
Height 
Upper Facial 
Index 
Spliakia ... 
Hawes 
140-6 
76-7 
54-56 
V. Luschan 
U3 
76 
54-39 . 
Selinos ... 
Ilawe.s 
140-9 
74-9 
53-20 
... ... 
V. Luschan 
143 
75 
52-51 
Pearson's criticisms are based on skull measurements and are more than borne 
out by Martin's table'-"" where the extreme range for the human race is given — by 
groups — as 49'0 — 56"0, but his remarks apply with still greater force to measure- 
ments on the living. 
Summarizing them there are two reasons for rejecting the upper facial index as 
a racial test, fii'st the possibility of error in measurements and secondly the small 
variation occurring in the human race. In regard to the first we have shewn that 
as far as can be judged from the limited data at our disposal independent observers 
have obtained similar results in Crete. In regard to the second the differences 
obtained in our area are sufficient to warrant a careful inquiry into the causes of 
this difference. 
We still have a marked distinction between our two big groups Crete and 
Cyprus, but the conditions of the smaller groups are different. Crete is in an 
intermediate position, a group of nearly 100 men from the Peloponnese — sufficient 
at any rate to point the way to the true condition in Mani — at one end and Cyprus 
at the other. 
Nasal Indices. 
The nasal indices (see Table IX) all shew as is usual a high degree of variation. 
They fall into two groups, the Meligalots, Maniots, Selinots and Sphakiots on the 
one hand and Cypriots on the other. The third Cyprian subgroup (Enkomi area) 
has a low index, the Leukadians occupy an intermediate position, the remainder are 
broader nosed. It would appear that the narrow-nosed groups are rather more hinno- 
geneous than the remainder, but the difference between the standard deviations, 
except in the case of Cyprus, is not sufficient to be of any significance. The fact 
that the means for Crete and Cyprus, each island taken as a single unit, are almost 
in exact agreement is striking, but the difference between the standard deviations 
