R. C. PUNNETT 
333 
Groivth Changes. 
Note was made above (p. 327) of certain marked discrepancies in the 
correlation values. 
(1) The correlation between anterior and posterior spines is much lower in 
embryos than in adults. Both anterior and posterior spines are rather more 
rostrally situated in the embryos than in the adults. The differences, however, in 
the (/s are small and will not serve to explain the very low correlation values of 
the embryos. Again the variability of the posterior spine is very close in all the 
groups. When, however, the variability of the anterior spine is examined it is at 
once evident that its value is much less for the embryos than for the adults. It 
would appear not unlikely that changes in the position of this spine occur during 
post-natal growth. As these changes are in the direction of higher correlation 
with a more variable structure, the posterior spine (as estimated by a), they must 
lead to greater variability in the anterior spine. Such is actually the case. 
Consequently we must suppose that the anterior spine is subject to a certain 
amount of change in position during growth. And in this connection it is 
interesting to notice that greater variability is associated with higher correlation 
also in the case of the characters involving the pelvic nerves of the cfs as compared 
with the ? s (cf Tables 2 and 30—35). 
(2) The correlation between the total segments and the half vertebrae for 
embryos is very low. The correlation between the total segments and whole 
vertebrae in this same group is however markedly higher than in the rest. Here 
again it is probable that growth changes come in. If during post-embryonic life 
we suppose that there is a certain amount of transformation of whole vertebrae 
into half vertebrae, such a process would lead to a reduction in the value of the 
correlation between whole vertebrae and total segments and an increase in the 
value of the correlation between total segments and half vertebrae. This is 
actually what has come about in the adult cTs. Moreover the correlation values 
bear out the supposition that the same process occurs, though to a less extent, 
among the $s. The correlation between the half vertebrae and total segments is 
less in the embryos than in the adults, whilst that between the whole vertebrae 
and the total segments is greater. It is therefore exceedingly probable that a 
certain amount of conversion of whole into half vertebrae occurs in either sex 
during post-natal existence. 
Variability in the Girdle Canals. 
That there may be either one or two canals in the pelvic girdle for the 
transmission of nerves has already been noted. The figures below seem to show 
without doubt that the number of the canals is connected with the position of the 
girdle. 
