204 BULLETIN OP THE UNIVERSITY OF WISCONSIN. 



have an entirely different type of epithelial cell, as shown in 

 PL 4, rig. 3. The cells are often extremely long and always 

 have very jagged edges, more so than the figure indicates. Be- 

 tween these types we find all gradations, though the last type 

 passes over into the others very abruptly, as a rule. 



In all my specimens I was not able to see a single spot or ori- 

 fice which I could construe to be a stomata. It is true that 

 there were some small spots in some of the preparations, but they 

 could always be traced to over distention, tearing, or drying of 

 the lung. Indeed, they could be produced in any quantity by 

 just such procedures. 



Along the inner side of the lung, near the pulmonary vein, 

 pigment cells are often found which present many bizarre 

 shapes. 



THE PULMONARY EPITHELIUM. 



If a Nedurus be killed by cutting off its head and one of its. 

 lungs be removed, cut open and spread out on a slide in salt solu- 

 tion, and its inner surface examined under a medium power, it 

 will be seen that the epithelium is apparently collected into 

 groups or islands of cells which are surrounded by clear spaces. 

 This arrangement coincides with that given by Schulze, Schmidt 

 and Ranvier for the frog, and by Williams and Stirling for the 

 newt. 



Schulze describes the relation between the capillaries and the 

 epithelium of the lung of Reptiles and Amphibians as follows : 



"All respiratory capillaries are attached to the alveolar wall only 

 on one side. They would project freely, with their greatest circum- 

 ference, into the air cavity, if they were not completely covered by 

 a continuous plate-like epithelium. The large polygonal cells of this 

 alveolar epithelium, meeting accurately at their sides, cover the sur- 

 face of the capillaries, turned toward the air cavity, with thin, trans- 

 parent, plate-like expansions and send plug-like continuations, gen- 

 erally the nucleus with some surrounding protoplasm, into the cap- 

 illary network and indeed so far down that they reach the connective 

 tissue stroma of the alveolar wall, and so completely fill the spaces 

 of the capillary network. These plug-like continuations, harboring 



