84 
A Study of the Crania of the Moriori 
results are given ili the footnote below * ; it suffices to remark here that for practical 
purposes it appears approximate enough if we deduct one-twentieth from the 
volume of the loosely packed seed in order to obtain the volume of the tightly 
packed. My corrected results from loose packing are given in brackets in the 
tables. The means are: 
Tight packing, S 1422-07 ; Loose packing, S 1423-03, 
$1310-73; „ „ ? 1308-47. 
The mean difference (regardless of sign) of the results obtained from loose 
packing from those obtained from tight packing is, however, 8 cm.^ How much of 
this is due to variation in the tight packing in the measuring glass, how much to 
fortuitous settlements in the loose packing it is not possible to say. The difierences 
suggest at any rate that a method may well be devised by which tight packing 
in the measuring cylinder may be avoided. 
Further measurements made are : 
F = Flower's Ophryo-occipital length. L' = glabellar horizontal length. 
L = maximum length from glabella to occiput. B = maximum horizontal 
breadth. B' = least forehead breadth from one temporal crest to the other. 
H = Basio-bregmatic height. OH = auricular height as measured on the 
* Let C( be the capacity from tight packing, C; from loose packing ; let <r be a standard deviation, 
V a coefficient of variation and the correlation of C( and C; , then in cubic centimetres : 
For male crania. 
C(= 1422-07 Mean C; = 1498-11 
= 110-185 (ri = 113-263 
Vt= 7-75 Vi= 7-66 
rti = -9935 
For female crania. 
Mean = 1310-74 Mean C, = 1377-37 
at = 90-265 ct; = 90-122 
Vt= 6-89 F(= 6-81 
r„ = -9969 
It will be seen that the correlations are very high. The regression lines giving for C; are : 
Ct = -96654 Gi - 25-91 for males, 
Ct = -99847 C, - 64-53 for females. 
We should have expected the same line to have resulted from the two series, as it does not it is clear 
that the coefficient of C; is a function of C; . To determine the nature of loose packing a little further 
we took the regression lines of (C; - C()/Cj and C, or equations of the form (C; - Cj)/Cj = ACi + B. 
There resulted : 
(C; - Ct)ICi = -06843 - -0000,1173 for males, 
(C( - Ct)ICi = -056939 - -0000,0619 (7, for females. 
These may be written : 
Gt = Ci (-93157 + -0000,1173 C,) for males, 
C( = Ci (-94306 + -0000,0619 C,) for females. 
The terms depending on C;^ are so small that we may insert in them for C;, the mean values in each 
case, when we find : 
Ct = Ci (-94914) for males, 
Ct = Gi (-95159) for females. 
The differences are now not unreasonable considering the paucity of the data, and we see that very 
approximately it suffices to take for both sexes, = C; (1 - vrV); the rule given above. 
The whole matter is to be reconsidered on more ample material. It is not clear at present how 
far it depends on personal equation in tight and loose packing of different amounts and how far on 
the diameter of the measuring glass, i.e. the pressure varying with height or with C"; and so affecting 
the looseness of the loose packing. 
