236 
EDWIN S. G0(3DR1CH. 
Further, the theory cannot afford even a formal explanation 
of the change of position of the median ffns relative to the 
paired fins in fishes, as I have already pointed out elsewhere 
(12). The median fins in Elasmobranchs are similar in 
structure and development to the paired fins, and, like these, 
shift up and down the body in different forms. Now the 
interesting and quite conclusive fact to notice is that the two 
sets of fins shift independently. For instance, the first dorsal 
fin is opposite the pectoral in Lamna, 'between the pectoral 
and the pelvic in Alopecias, opposite the pelvic in Scyllium, 
and behind the pelvic in Raja. If it is granted that the fins are 
homologous in these four genera, no addition or suppression 
of segments can possibly account for their disposition. Other 
cases of homologous organs passing in front or behind one 
another could be given. 
The evidence against the theory of intercalation is over- 
whelming. Moreover, it undermines the very foundation of 
the definition of homology it is intended to uphold ; for if 
segments can be added to or removed from any part of the 
series, there is no guarantee that any one particular segment 
in one individual really corresponds to any particular segment 
in another individual. 
Let us now briefly examine a view which may be called the 
theory of re-division. It may be stated as follows : if one 
individual is composed of, say, twenty segments, and another 
of twenty-one,'the difference is due, not to the addition of a new 
segment, but to the subdivision of the individual in the first 
case into twenty segments and in the second case into twenty- 
one segments. Therefore, no segment of the first individual 
can be strictly homologous with any segment of the second. If 
the number of segments is sufficiently increased by further sub- 
division into twenty-two, twenty-three, and soon, an organ orig- 
inally situated on, say, the fifth segment might later on be found 
on the sixth or seventh ; and two organs originally close to- 
gether might become separated by a larger and larger number 
of segments. Although this theory may appear promising at 
first sight, it soon becomes obvious that it could only apply to the 
