XXXil INTRODUCTION. 
of scales and fin-rays, but the latter appears to have constantly 8 gill-rakers on the 
lower part of the anterior arch, whilst the former has 7. 
Centropomus viridis and C. undecimalis agree in the number of scales and gill-rakers, 
but the former appears to have a ray less in the second dorsal fin. A similar difference 
is found between Philypnus maculatus and P. dormitator, the former having a ray 
more in the anal fin. 
The members of each pair of species in all cases differ from each other in at least one 
more or less tangible “ substantive” character, such as the relative depth of the body, 
length of the head, size of the mouth, length of the fin-rays, or size of the teeth. 
When there is agreement in all the characters which can be numerically expressed 
and the distinction appears to rest on a single feature, as in the case of Dormitator 
latifrons, separated from D. maculatus by the larger head, or Gerres lineatus, differing 
from G. plumieri in the less elevated dorsal fin, the two forms can scarcely be regarded 
as more than sub-specifically distinct. 
In other cases the Atlantic and Pacific forms differ in two tangible characters, 
e.g. Gerres olithostomus and G. peruvianus, the latter of which is distinguished by the 
larger scales in the thoracic region and the somewhat smaller mouth. From such 
pairs of species we pass to others, the members of which may be separated by three, 
four, or more tangible characters. 
From the data given above we may conclude that a greater or less amount of change 
may take place in a species when isolated for a considerable period, and that some of 
the changes which take place may be non-adaptive. 
The fact that in so many cases species on opposite coasts may be paired is more in 
harmony with the view that there has been a gradual modification during isolation 
than with the supposition that a “mutant” has arisen which has replaced the 
parent form. 
C. T. R. 
February 1908. 
