THOEACIC GLANDS IN THE LAEViE OP TRICHOPTEEA. 411 



relationship. But this is not the case : nothing is known of the 

 development of these till lately undiscovered organs. And if they 

 were known to be epiblastic, as they probably are, it would not 

 settle the question, as they could then be the remains of the outer 

 part of the nephridia which so often originates as an epiblastic in- 

 growth, the mesoblastic or proper nephridial part having vanished 

 in the course of evolution. No conclusion could be drawn asfainst 

 their nephridial relationships, whatever might be their origia. 

 They appear to me, however, to be more likely nephridial than 

 coxal, for the following reasons : — 



1. They have no couuection with the appendages. This fact, 

 though not finally disposing of belief in their coxal nature, seems 

 worth consideration, as no organ undoichtedly coxal is known 

 to have moved far from the limb and met its fellow in the median 

 line. 



2. On the other hand, certain organs, the nephridial sig- 

 nificance of which it is scarcely possible to doubt, unite in the 

 median line and open there through one common aperture. Such 

 are the so-called " salivary glands" of Peripatus. These are long 

 tubes entirely disconnected in the embryo, and provided each 

 with a funnel or nephrostome. Later on they lose their inner 

 opening, and meet at the median line, just as the thoracic glands 

 do in Trichoptera. The same is true of the disposition of the silk- 

 glands of larval insects and, in many an adult form, of the true 

 salivary glands, both being considered as modified nephridia. 



3. There is a striking analogy between the arrangement of 

 the tubules of the thoracic glands of Trichoptera and that of 

 the Malpighian vessels generally. Both are derivatives of two 

 chief tubes (at least this is the primitive disposition of the Mal- 

 pighian vessels). These chief canals open in both cases through 

 a single epiblastic ingrowth, and the common duct of the thoracic 

 glands would thus appear to be equivalent in its relationships to 

 the proctodseum. "We have now much reason for regarding the 

 Malpighian vessels as modified nephridia ; and Gegenbaur's hypo- 

 thesis that these vessels primitively opened on the surface of the 

 body has received a strong confirmation from the fact, discovered 

 by Wheeler *, that in Doryphora they early appear in the form 

 of ingrowths from the walls of the proctodseum, while this 



* Wheeler, " The Embryology of Blatta germanica and Doryphora decem- 

 lineata" ' Journal of Morphology,' toI. iii. 1889. 



