152 MK. E. J. TILLYAED ON THE EECTAL BEEATHING-APPARATDS 



line in a slanting direction, so that its rounded and free distal end comes to. 

 overlap a corresponding gill from the next consecutive hemibranch situated 

 slightly more posteriad. The two sets of overlapping gills belong to one com- 

 plete holobranch. Further, each gill is in the form of an elongate-oval leaf on 

 a broad base, with its upper surface very concave, giving it almost a grooved 

 appearance. Also, the free border of each leaf is waved or undulated, some- 

 times only moderately as in Austroceschna, sometimes very deeply as in 

 Dendroceschna. In the latter case, the complicated appearance of the newly 

 everted gill-basket almost defies description. 



Dr. Ris's short account of the gills in Bracliytron show that they resemble 

 very closely those of AustrowscJma. He says that they are "leaves which 

 project like tiles one over another" ("Blatter, die sich dachziegelformig 

 llberlagern"). 



There are some important differences to be noted in the gills of the four 

 genera here under consideration. In Dendrocesclma and Austroplilehia there 

 is no sign either of fat-globules or of pigment, and the whole apparatus is of 

 a pure shining white colour. The frills along the free borders of the gills 

 are very numerous and complicated, especially in Dendroceschna, where the 

 whole oill-formation resembles nothing so much as a turbulent sea of waves 

 and wavelets inextricably intermingled. Transverse sections of the rectum 

 of this larva reveal the almost unbelievable extent to which this complex 

 folding and frilling is carried. 



In the genera Austroceschna and Brachytron there is no complicated 

 frilling of the edges of the gills. They are, instead, fairly regularly waved 

 and undulated. There is a considerable amount of pigmentation of a purplish- 

 brown colour. 



It is interesting to note that in the genus Austroceschna there is a distinct 

 tendency shown, in the posterior region of the gill-basket, towards a con- 

 striction of the broad bases of the gills. Thus the gills in this part of the 

 rectum tend to approach the Foliate Type as developed in the genus y^schna 

 (p. 153). If we now also bear in mind the fact that it is just in this posterior 

 reo-ion of the gill-basket, in yEschna itself, where the gill-folise are most 

 widely separated off and most definitely developed, we may well claim to 

 have here definite evidence that the earliest form of Implicate Type gave 

 orio-in to the Foliate Type during the course of evolution. 



Arrangement of the Trachece. — Each separate gill is supplied by a strong 

 thick secondary efferent trachea, which approaches its base from the side 

 nearest to the middle line of the holobranch to which it belongs. This 

 secondary efferent bends sharply down into the base of the gill, branching 

 into several stout gill-trachese. These run straight up towards the distal 

 border of the gill, giving off numerous capillaries and also smaller branches 

 which also quickly break up into capillaries. The structure of the free 

 border and the arrangement of the capillaries resemble very closely those 



