IN THE LARV^ OF ANISOPTERID DEAGONFLIES. 171 



of this organ in other forms besides Lihellula ; for, as we shall show, the 

 basal pad was first formed much later tlian the establishment of the syncytial 

 nature of the rectal epithelium, and hence must be regarded as never having 

 possessed separate cell-territories at all. 



Looking through our series of gill-forms, we are at once struck with the 

 fact that it is not in the Simplex System, which is phylogenetically the most 

 archaic, that the simplest form of basal pad is found. This is, perhaps, 

 explicable on the supposition that the more isolated positions of the gill-folds 

 in the Simplex System render the support of basal pads more necessary than 

 is the case with the Duplex System. In any case, it is a fact that well- 

 developed basal pads occur in all forms except the Implicate Type of the 

 Duplex System. We shall therefore select, first of all, for study the basal 

 pads, if such they may be called, of Dendroceschna conspersa ftext-fig. 16, hp). 

 It is only necessary to recall, in attempting to explain the rudimentary nature 

 of these structures, the fact that in the Implicate Type of gill the separate 

 grooved folds lie alternately with their rounded ends supported along a 

 wavy central line, so that there is actually no complete separation of the 

 separate portions of the holobranch into folise or lamellae. Hence the holo- 

 branch is more or less self-supporting, and a well-developed basal pad is not 

 wanted. 



In Dendrocesclina the base of the gill-fold is only very slightly enlarged 

 to receive a narrow elongated strip of hypobranchial tissue (%), in which 

 the gill-trachea {tr) runs so straight and regularly that it is very often met 

 with from end to end in the same cross-section. Our figure shows it cut in 

 two places. On the inner side of the gill-fold, the ordinary flat syncytial 

 epithelium borders the hypobranchial tissue along its whole length, and 

 passes round along an arc of the rectal circumference for a short distance, to 

 rise again towards the interior on the internal side of the next gill-fold. 

 But, in sections which show the efferent trachea, the external epithelium of 

 the base of the gill- fold is slightly swollen into a rudimentary basal pad {hp) . 

 The protoplasm of this pad shows, so far as I have been able to study it, no 

 trace of any fibrillar structure. It is, in fact, similar to the protoplasm of 

 the epithelial svncytium in its ordinary flat condition, except that it stains 

 slightly darker with hsematoxylin. Towards the two ends of the pad the 

 nuclei are rather numerous, but not different in size or structure from those 

 met with in other parts of the syncytium. But, in the main portion of the 

 pad, the nuclei are very distinctly enlarged, mostly very distinctly oval in 

 form, and show granular contents, and in many cases a fairly distinct 

 nucleolus. On the whole, these nuclei stain less deeply than those of the 

 syncytium, and their contents are more clearly differentiated. 



Passing away from the level of the efferent trachea, the sections show a 

 rapid decrease in the width of the basal pad, which very soon merges into 

 tho typical epithelial syncytium of the giil-fold. These pads, therefore, 



