368 On the Morphology of the Semicircular Canals. (June, 
and see how much of general application can be gleaned from 
the facts. In teleost fishes the side organs first exist as mere cel- 
lular elevations! projecting into the water and having the percipient 
structures, the hairs, protected (generally, not always) by means of 
a delicate hyaline cylinder open at its distal end. In most cases 
this stage is transitory. Above and below the elevations, the 
epithelium of the surrounding skin becomes raised, forming a 
groove which is finally converted into a canal. Thus these eleva- 
tions or macula of the side organs become enclosed by a process 
somewhat analogous to the involution of the ear. 
In Elasmobranch fishes, however, another method is pursued. 
The lateral line appears already in the embryo as a fully formed 
canal, and, instead of being produced by an inflexion of the 
epiblast, is formed directly by being hollowed out in the substance 
of the same.” This departure from the general method of the 
development of sensory structures is an illustration of an import- 
ant fact dwelt on by Herbert Spencer, namely, that the processes 
pursued in the development of the embryo are by no means the 
exact repetition of those which occurred in the evolution of the 
race. In many respects the processes of embryology are short- 
ened. The integral parts or molecules of many structures tend 
to assume directly the relations which they occupy in the adult 
form, without passing through those intermediate stages which 
were successively traversed by the ancestors of the races. : 
This marked difference between the development of the lateral 
lines of Elasmobranchs and Teleosts is in harmony with several 
other interesting facts. We would naturally suppose that an or- 
ganism in which a comparatively complex structure is produced 
by the direct method, is relatively both older and higher than one 
in which the same organ is produced by the indirect method. In 
the present instance this is the case. The Elasmobranchs consti- 
tute an order of fishes of vast antiquity, making their appearance 
already in the Upper Silurian, and being, therefore, among the 
oldest fishes known. Again, they constitute a very high order 
being perhaps, with the exception of the Dipnoi, the highest. 
They are even related in some respects to amphibians. The 
Teleosts, on the other hand, are ordinarily regarded as the repre- 
sentatives of the piscine type ; that is, as presenting preéminently 
those characters which constitute a fish. Secondly, the Teleosts 
1 The nervehills of F. E. Schulze. 
= 2 See Balfour’s Monograph of the Development of Elasmobranch Fishes, 1877- 
