OEIGIN OF METAMERIC SEGMENTATION. 



65 



derived from its walls. This may reasonably be explained in 

 the same way as I have attempted to explain in my paper 

 quoted above, the development of the hinder part of the 

 Amphibian kidney (modified larval development). 



-4j$ 





Fig. 1. 



-Diagram illustrating the development of the excretory system of 

 Polygordius (from Balfour, after Hatscliek). 



The development of the excretory organs in Polygordius 

 (woodcut, fig. 1) as described by Hatschek, is explicable on my 

 hypothesis and so is confirmatory. The temporary longitu- 

 dinal canal^ which at first connects all the organs, is obviously a 

 rudiment of the longitudinal duct found in the Vertebrata. 

 The presence of this duct indicates that in the diploblastic 

 ancestor of Polygordius, a system of canals was present in the 

 coelom together with the excretory pores. 



On the Origin op Trachea and Gill Slits. 



The view that tracheae are derived from the cutaneous 

 glands of a worm-like ancestor with a well developed middle 

 layer is beset with so many physiological difficulties that I 

 venture to suggest the following hypothesis, which agrees 

 equally well with what we know of the development of 

 tracheae. 



Trachete had their origin, like the organs so far discussed, in 

 the diploblastic ancestor. In this ancestor they had the form 



VOL. XXIV. NEW SER, E 



