350 The Philippine Journal of Science i»23 



is greatly increasing our knowledge of them and making clear 

 their affinities. 



Any arrangement based upon the position of the ocelli, used 

 as quite apart from the structural modifications and accom- 

 panying sculptural features of the cephalic sclerites that bring 

 about the apparently different positions of the ocelli, can only 

 lead to utter confusion when an attempt is made to arrange 

 the vast tropical faunee by this ancient and impossible "system." 

 If, instead, we use the fundamental structure of the head, we 

 quickly associate what are evidently closely related forms that 

 were formerly scattered through various sections of the old 

 system. Out of the remarkably homogeneous series of species 

 in the Nirvaniidae, with a remarkable similarity in fundamental 

 anatomy and likeness in bodily form and even in color patterns, 

 some would be placed in Phrynomorpharia, some in Acocephalini, 

 and some in Tettigoniellidae — and this has actually occurred. 

 The following tentative synopsis will bring out some of the 

 relationships discussed above. It does not presume to be the 

 presentation of a complete system for the Jassoidea. Some of 

 the most important anatomical features of very many jassoid 

 genera are undescribed and omitted from figures, so that a 

 general system could not be developed without visits to many 

 of the European collections. The new families proposed are 

 even more distinct than Ulopidse and Paropiidse, which have 

 been long recognized as good primary groups. The "Jassidse" 

 of this synopsis will have to be further divided. Some of the 

 divisions long recognized, like the Eupterygini, have never been 

 diagnostically characterized. The accepted characterization of 

 Eupterygini will certainly as readily admit to it many tropical 

 forms of other groups like Nirvana that have no relation to 

 true Eupterygini. 



In future studies and especially in reexamination of previous 

 types a careful comparative study of leg structure seems certain 

 to yield results of great value. The great diversity of structure 

 in the hind tarsi is shown in the accompanying figure (Plate 1, 

 fig. 1). In Onukm (Plate 1, fig. 1, a) and Pythamus (Plate 1, 

 fig. 1, b), of the Pythamida?, the third joint is inserted in 

 second, and second in first, far before the apex, and the length 

 of the first joint is as great as, or greater than, the following 

 two together, the first and second joints being crowned with 

 groups of stout spines or teeth. In both cases there are no 

 lateral spines on the first joint except at the insertion of the 

 succeeding joint. In Stenometopius (Nirvaniidse-Stenometopii- 



