On Amphibian Metamorphosis and Internal Secretions. 47 



in other Urodeles, indicated the advisability of extending such observations. 

 It was not possible to obtain specimens of Proteus ; but, through the 

 courtesy of Prof. MacBride, three live medium -sized Necturus were secured^ 

 and submitted to thyroid treatment in his laboratory at the Imperial College 

 of Science. 



The experiment began on November 10, 1920. One animal was kept as a 

 control, and fed on small pieces of raw beef : the other two were given fresh 

 thyroid (ox) gland tri-weekly, as in the case of the first experiments with 

 Axolotls. Up till the time of writing, the treatment has been continued 

 without interruption for 7 months. No pigmental changes have resulted ; 

 and there has been in neither case any appreciable reduction of the tail-fin. 

 As regards the condition of the external gills, observation is embarrassed by 

 the fact that the filaments are in a very marked degree erectile, their 

 length, when fully dilated with blood, being many times greater than 

 when the animal is not actively respiring : they can, however, be induced to. 

 extend by compelling the animal to perform muscular exercise, after which 

 a rough estimate may be made of their maximum dimensions. Constant 

 attention to this point showed very clearly that the filaments of the experi- 

 mental individuals were relatively shorter even when fully extended ; but, 

 for. an obvious reason, it is not possible to interpret this as necessarily 

 consequent upon a reduction in the actual amount of tissue ; for it is well 

 known that thyroid administration influences the blood pressure in Verte- 

 brates, and the behaviour of the gills in Necturus is evidently a vaso-motor 

 phenomenon. 



It does not seem likely, therefore, that an administration of thyroid in this 

 Perennibranchiate form is effective in producing somatic modifications com- 

 parable to those occurring at metamorphosis in other Urodeles. In view 

 of the morphological data given by Piatt, and of the experimental evidence 

 available, it may therefore be stated that, if the Perennibranchiate forms 

 like Necturus and Siren are not primitive in the invariable retention of 

 the larval type of Urodele organisation throughout life, their failure to 

 develop the predominantly adult characteristics is not due primarily to 

 thyroid deficiency — this, of course, does not apply to Typhlomolge, and 

 possibly to Proteus also — nor to a defective supply of iodine in the 

 environment. There remain at least four possible interpretations, then, of 

 the Perennibranchiate state : — 



(i) That these animals have never possessed genetic factors responsible for 

 the structures of typical adult Urodela ; i.e., that in this respect they are 

 actually primitive. This appears to be negatived by the purely morpho- 

 logical evidence available (see Gadow, 1901, pp. 65, 136). 



