114 MR. G. W. SMITH AND DR. E. H. J. SCHUSTER ON 



else that the two species of Engceus have been somehow trans- 

 ported across Bass' Straits by accidental means. 



Now, it transpires as tlie result of a detailed examination of the 

 various species of Engceus, especially in respect to the structure of 

 their gills, that the derivation of Engceus from ParachcerajJS is 

 entirely false, and that the superficial resemblance of these forms 

 to one another is due to convergence. On the other hand, it 

 is abundantly evident from the gill characters that Eiigceus is 

 a derivative of Astacopsis, and has nothing to do with either 

 Chceraps or Parachceraps, The reasons xipon which this conclusion 

 is based are as follows. In Astacopsis the podobi-anchia do not 

 possess a broad ala or wing-like expansion of the stem of the gill ; 

 the ala is, on the contrary, reduced to a mere rudiment, as shown 

 in the ti-ansverse section (PL XII. fig. 1). Attached to this 

 rudimentary ala of the podobranchs are a few gill-filaments, 

 which are furnished at their tips with charactei^istically shaped 

 hooks (fig. 2). All the other gills in Astacopsis, exclusive of the 

 podobranchs, u e. the arthrobranchs and pleurobranchs, have their 

 filaments entirely free from terminal hooks. The peculiar hooked 

 setae present on the bases of the podobranchs in Astacoj)sis also 

 have a constant and characteristic shape, the terminal hooks 

 being not sharply recurved (fig. 3). Now all the above characters 

 are absolutely constant for the various species of Astacopsis. In 

 Chceraps and Parach/xrdps, on the other hand, we have a totally 

 different series of gill charactei-s which are just as constant and 

 characteristic for all the species of these two closely related 

 genera. In these forms the podobranchia possess a very broad ala 

 (fig. 4) which is furnished with numerous filaments, whose hooks 

 have a characteristic sickle shape (fig. 5). In all the other gills, 

 besides the podobranchs, numerous filaments are provided with 

 these terminal hooks. Finally, the hooked sette (fig. 6) on the 

 podobranchs have an entirely difierent shape compared with those 

 of Astacopsis, being sharply recurved at the ends. 



Now, the distinctive character and absolute constancy of these 

 gill-structures force us to attach especial taxonomic importance 

 to them, far more importance than external appearance or other 

 characters which fluctuate from species to species, so tha.t an 

 examination of the gills of Engmus should give us the key to its 

 relationship with the other genera, "the gills o/Engjeus ago^ee in 

 all the above particulars ivith those o/ Astacopsis, and differ entirely 

 from those of either Chseraps or Parachseraps (PI. XIII. figs. 7-10). 

 Thus the podobranchs possess a rudimentary ala (figs. 7, 8) which 

 carries a few filaments, and these filaments terminate in hooks 

 (fig. 9) shaped like those of Astacox>sis. The gills other than 

 the podobranchs have their filaments unprovided with hooks; 

 the hooked sette (fig. 10) on the podobranchs are not sharply 

 recurved at the end, but are shaped as in AstdcOpsis, 



Now, apart from the gills, Engceus diflers so widely from both 

 Astacojjsis, Chceraps, and Parctchceraps in the characters of its 

 bodily structure, appendages, etc., that these no longer serve as a 



