244 
EDWIN fS. GOODRICH. 
Teleostean family Triacanthidae is composed of twenty verte- 
bras, and so on. Undoubtedly a correspondence of this kind 
may become establitshed, but it appears to be of secondary im- 
portance. Generally speaking it is more definite and invariable 
in the anterior than in the posterior region, and in regions or 
animals composed of few than in those composed of many 
segments. It is just as if Nature got tired of counting towards 
the tail end of a developing animal, and as if her arithmetic 
became uncertain when dealing with large numbers. But 
structure and segmentation vary independently, and whatever 
may be the connection which becomes established between 
them, and however close it may be, it would seem that we 
must not consider it as constant and essential. 
In the segmented vertebrate the materials for the formation 
of muscular nervous and skeletal segments are distributed 
along the body ; those particular segments which occur in ap- 
propriate positions are entrusted, so to speak, with the develop- 
ment of special organs. The function of producing an organ 
may be transposed from one segment to another. How this 
transposition is brought about we do not know as yet; pos- 
sibly it is accompanied by the redistribution of organ-forming 
substances at a very early stage, regulated by a complex of 
stimuli subordinated to the needs of the individual as a whole. ^ 
^ What has been said of the Vertebrates would probably apply equally 
well to the Annelids or Arthropods. Although' their segmentation may 
very well have been quite independently developed, yet it resembles 
that of the vertebrate in that segments appear to be always increased 
or diminished in number at the posterior end of the series. The same 
specialisation of segments, of appendages, and of other repeated parts 
takes place, and the same apparent shifting. In these also it is to be 
explained by transposition. Any appropriate segment in the series 
may be called upon to develop a particular kind of appendage or a 
genital duct, for instance, and the number of these organs may be 
increased or diminished without a corresponding change in the whole 
number of segments. We may speak of a general homology of the 
sevei*al series of repeated parts, and of a more special homology of any 
one of them which, like the mandible or green gland or oviduct, may be 
traced to a common ancestor. Just as in the vertebrate, so in the 
annelid and arthropod, a very constant relation may be established. 
