An endeavour to show that the tracheae arose from setiparous sacs. 517 



cause it has been clearly shown that the salivary glands of Peripatus 

 are a pair of nephridia brought into the buccal cavity by the de- 

 velopment of a kind of rampart which encloses their openings with the 

 mouth and jaws. This very remarkable process does not seem to 

 have any parallel in the development of the Hexapoda; it is never- 

 theless possible that the salivary glands may be nephridia, as is some- 

 times assumed. I do not, however, think that this origin is so likely 

 as the one here proposed, which homologises the salivary glands, 

 larval sericteries and tracheii) with a row of chsotopodan acicular glands. 

 Not only are the salivary glands of the Insecta of purely ectodermal origin, 

 but the fact that the process of the conversion of one pair of neph- 

 ridia is visible in the development of Peripatus shows that this arrange- 

 ment of salivary glands was comparatively recently acquired in this 

 particular case, just as, also in this particular case, the acicular gland 

 of the 2nd pair of appendages (excluding the antenna^ develop into 

 slime or spinning glands, whereas in the Hexapoda they appear to develop 

 into salivary glands. This of course assumes that each pair of sali- 

 vary glands belongs to a different segment^). 



Perhaps the homologies here claimed may help to solve some of 

 the vexed questions with regard to the morphology of these salivary 

 glands. In the meantime I think there is some reason for believing 

 that they and the tracheae are to be referred to acicular glands. 



Protracheata. 



The development of the tracheae from setiparous glands, again, 

 throws light on their diffuse arrangement of the former in the Pro- 

 tracheata, as compared with their strictly metameric arrangement in 

 the Hexapoda. 



This remarkable diflerence in the arrangement of the tracheae has 

 always prevented the establishment of any very close relationship 

 between Peripatus and the Hexapoda. It is true that some species 

 of Peripatus (e. g. P. capensis) show a tendency to develop rows of 

 stigmata, but this can easily be explained on the theory I am here 

 trying to establish. As above stated, we can deduce the simple ecto- 

 dermal invaginations which, we think, were the starting point of 

 tracheae, not only from acicular glands, but from any setiparous glands. 



1) cf. KoBSCHELT und Hbideb, Lehrbuch dei- Entwicklungsgeschichte, 

 p. 832. 



