GENERAL MORPHOLOGY. 357 



These facts led Palmen to the conclusion that the spiracles 

 are virtually, although not functionally, present in the larva, 

 and that the closed system is a modification of the open 

 system. This view does not necessarily follow, however, from 

 the facts, for if the tracheae are developed from parablast, the 

 existence of cell-strings uniting the tracheal trunks and the 

 skin might be expected ; and the manner in which ecdysis 

 takes place may be regarded as an adaptive modification. 



The ecdysis of the intima of the tracheae in the fly nymph 

 occurs, nearly in the same way, as in larvae with closed tracheae ; 

 new spiracles are formed, which are quite independent of the 

 larval spiracles, whilst the vestibule of the spiracle and the valves 

 are developed from the integument, and not from the cell-string, 

 through which the intima of the tracheae of the pronymph is 

 withdrawn (see PI. XXII., Figs. 6 and 7). 



Palmen further regarded the longitudinal trunks as originating 

 by the anastomosis of tracheae springing directly from the 

 spiracular sacs. In the higher insects this does not appear 

 to be their origin, and the existence of these vessels is so 

 constant in all insects, whilst the position and number of the 

 spiracles is so variable, that it can hardly be doubted, I think, 

 that they originated in Tracheata in which the spiracular vessels 

 did not exist. 



The only instance in which the longitudinal trunks are 

 absent in the less highly differentiated Tracheata are the highly 

 abnormal Protracheate, Peripatus, and in some Thysanuridae. 

 In Peripatus the tracheal vessels are so irregular that they 

 cannot be taken as the original form of the very constant 

 tracheal system of the Myriapoda and Insecta, and it appears 

 equally probable that the genus Peripatus is a highly degene- 

 rate form. With regard to the condition in some Thysanuridae, 

 described by Palmen, it is quite as possible that the tracheal 

 system has undergone retrograde metamorphosis. The Spring- 

 tails cannot be regarded as generalised Orthoptera, and there 

 is no evidence that retrograde follow the same lines as 

 evolutionary modifications. 



It appears to me most probable that the tracheal system is 



