234 
EDWIN S. GOODUICIT. 
ap|)roximatioii of tlie pelvic fin to the ])ectoral had been 
brought about by the excalation of segments, we should have 
to assume that the whole trunk region between the two had 
been suppressed in Gadus and Torpedo — an obviously ai)surd 
assumption, as Fiirbringer has already pointed out. Moreover, 
twenty-seven new segments would have to be intercalated in 
the region of the pectoral fin of Torpedo. Now since 
such vast differences frequently occur between closely ^dlied 
forms, w’e should have to suppose that this wholesale 
appearance or disappearance of segments takes place rapidly 
and repeatedly. If such were really the case, shonld we not 
ex])ect to find some indication of the process — zones of growth 
or zones of reduction, where new segments would be forming, 
or old ones vanishing? No trace of such zones is found, 
either in the adult or in the embryo.^ On the contrary, 
segments are found to grow only at the extreme end of 
^ It may, of course, be suggested that segments can l^e intercalated 
and excalated in the course of phylogeny just as they appear to be in 
ontogeny. But the two processes must be clearly distinguished. First 
of all there is, I believe, no evidence that segments are ever intercalated 
in ontogeny. Observation seems to have established that new segments 
are only formed at the growing end of the series. On the other hand, 
it is obvious that segments may be more or less completely crushed out 
in ontogeny — as, for instance, in the occipital region (see below). But 
it must be remembered, however, that these segments are, as a rule, 
clearly formed in early stages of development, and that our inability to 
make them out in later stages is mainly due to practical difficulties of 
observation and technique. The work is done on a series of individuals 
killed at various stages of development ; often the series is incomplete 
and the stages separated by wide gaps. Could we observe one 
individual continuously, we should probably be al^le to trace out the 
fate of each segment with certainty. As a matter of fact the somites 
of the occipital region are generally distinct enough in the early stages 
of the most modified Craniates. Still, it may be supposed that if the 
X>rocess of reduction began earlier and earlier in successive generations, 
the crushing out might at last take place at the very beginning of 
growth, and that these segments might finally cease to appear at all as 
such. It would be rash to say that this has never hapx)ened ; but, so 
far as I know, there is no convincing evidence that it has ever been the 
fate of segments in the Vertebrata. 
