TRANSFORMATION OF EPITHELIUM. 521 
in another, the same spot, perhaps protected in some way from 
friction, is covered by ciliated epithelium without cilia. This 
suggested to us that during the life of the animal the mem- 
brane, at first entirely covered by cilia, became modified at those 
spots, and at those spots only, where friction actually occurs. 
In order to settle this point sections were prepared from an 
embryo kitten five centimetres in length, by embedding in 
celloidin in the usual way, and carefully examined. The car- 
tilage ring and trachealis muscle were both present. The 
former was, however, incomplete, and did not extend round 
more than three fourths of the tube. Where the cartilage 
was absent the mucous membrane was thrown into well- 
marked folds. The whole cavity was lined by stratified ciliated 
epithelium, the superficial cells of which were vertical to the 
surface. This character was uniformly preserved in all the 
specimens examined. 
Sections of a young kitten about two weeks old were then 
prepared. The cartilage rings had met, but they had not over- 
lapped. One end was evidently turning inwards, and was 
lying in the direction of a projecting ridge of mucous mem- 
brane. The epithelium lining the cavity was ciliated through- 
out its whole extent (fig. 6). 
It appears, therefore, that the transitional and stratified 
squamous epithelium present in the adult trachea are produced 
from a fully differentiated ciliated membrane. 
In this case it is probable that as a result of the action of 
environment—friction with an opposing surface—the ciliated 
cells were gradually destroyed, and that the more deeply placed 
and less differentiated cells were induced to take on another 
line of development. They were modified into cells, which, by 
their division, no longer produced elongated and ciliated epithe- 
lium, but were capable of forming flattened keratinized cells. 
Still more remarkable changes, from one cell structure to 
another, have been discovered by biologists. Of Hunter’s 
gull we believe there is no histological record; but we know 
that an endothelial cell may come to resemble a white blood- 
corpuscle, and the latter may, in its turn, change into a con- 
