An endeavour to show that the tracheae arose from setiparous sacs. 519 



These lungs project into the pericardium and aerate a blood stream 0, 

 thus offering an interesting parallel to the lung books of the Spiders 

 which also aerate a blood stream, whereas the tubular trachea; seek 

 out the tissues themselves and oxygenate them directly. The lungs of 

 Scutigera may be developed from groups of ordinary bristle sacs which, 

 as we have seen , appear to have formed the tracheae in Peripatus. 

 But there can be no doubt that in this, as in many other respects, 

 the Scutigeridce are highly specialised. 



The most primitive condition, among the other Myriapoda at least, 

 is found among the Diplopoda, in which the tracheae are metameri- 

 cally arranged as in the Hexapodan embryo, and may then be homo- 

 logised with the acicular glands of the dorsal parapodia. It is an 

 important point for our argument, that on every double segment there 

 are two stigmata, showing thereby that a very intimate connection 

 exists between the limbs (or parapodia) and the tracheae, a connection 

 which our theory fully explains. The stink glands which occur in 

 each segment in the lulidce just above these acicular tracheae may 

 perhaps be the glands of the original parapodial setae. 



The presence of coxal glands in the Myriapoda, also deducible 

 from setiparous (probably acicular) sacs, is especially interesting, 

 leading us on to the last group which we have to consider — 

 the Arachnida. 



The Arachnida. 



An attempt has been made to deduce the tracheae of the Arach- 

 nids from the abdominal gills of a Limulus-Wke ancestor. At the 

 outset, we repeat what we have said with regard to the nephridial 

 origin of the tracheae "it will not explain all the conditions" ; it leaves 

 the cephalothoracic tracheae of the Acarids and of the Solpugidœ to 

 be separately accounted for. Our theory avoids this difficulty. 



On turning to the Arachnida we naturally expect diff"erent con- 

 ditions, considering the specialisation of this group. Their whole or- 

 ganisation shows them to have been very early diff'erentiated from the 

 primitive Tracheatan-Annelid. It is however a striking fact that the 

 row of structures corresponding with the Hexapodan spinning and 

 salivary glands and tracheie is not lateral but ventral. We have the 

 poison glands in the chelicerae, the maxillary glands in the basal joint 



1) see Sinclair, A new mode of respiration in the Myriapoda, in 

 Ann. and Mag. N. Hist. March 1892. 



