An endeavour to show that the tracheœ arose from setiparous sacs. 515 



which are but folds of the skin. It seems to require uo great stretch 

 of the imagination to believe that, as the dorsal parapodia gradually 

 went out of use, owing to the specialisation of the ventral parapodia 

 as ambulatory limbs in adaptation to a land life, the acicular glands 

 might acquire a new function. They would obviously be no longer 

 required to secrete the large supporting bristles ; they would therefore, 

 if we may use the expression , be at the disposal of the animal for 

 use in some other way. It would, further, be the acicular glands in 

 preference to any other setiparous glands which would be thus liable 

 to modification; because the more protective bristles would probably 

 be as functional on land as in the water. 



We have, indeed, among the Chaetopod Annelids, an almost exact 

 parallel to the case here imagined. In Polyodontes (a Chaetopod An- 

 nelid) the dorsal branch of the parapodium degenerates and fuses with 

 the ventral branch. Its acicular gland is retained as an additional 

 skeletal support for the fused parapodia. But its bristle glands develop 

 into what Eisig ^) calls a "spinning gland". This fact is especially 

 interesting to us because, as we shall see below, these bristle- and 

 acicular glands in the Tracheatan-Annelid not only formed tracheae 

 but also spinning glands. 



Reference to Figs. 2 and 3 shows how well this origin of the tra- 

 cheae agrees with facts. We find in each trunk segment dorsally to 

 the limb, i. e. exactly in the place of the vanished parapodium (as- 

 suming the limb to represent the ventral branch), a stigma which 

 seems to me to receive its full explanation as a derivative of an aci- 

 cular gland. 



These tracheae are, however, not the only developments of the 

 acicular glands in the Hexapoda. There are other structures most 

 important for our argument. BiJtschli 2) has already claimed the 

 larval spinning gland on the 2nd maxilla as homologous with the tra- 

 cheae, i. e. as the modified tracheal invagination of the 1st trunk seg- 

 ment. This, as I have just stated, is most important for our argument, 

 because there can be little doubt that spinning glands are, as a rule, 

 to be homologised with setiparous glands, such a transformation oc- 

 curring, as above described, among the Annelids. If, then, this larval 

 spinning gland is homologous with the tracheae, it leaves little doubt 



1) cf. Die Capitelliden des Golfes von Neapel, p. 324. 



2) Zur Entwicklungsgeschichte der Biene, in: Zeitschr. wiss. ZooL, 

 Bd. 20, 1870. 



