432 FAUNA HAWAIIENSIS 



articulation with the trochanter occurs, is externally visible ; all the rest is completely 

 covered in by the metasternum and first abdominal segment, which have met above it ; 

 the normal articulation of the epimeron with the outer part of the coxa is present, but 

 quite concealed from view. 



By dissection of several specimens of the Fallia from the Hawaiian Islands, I have 

 found that exactly the same condition exists in that genus. 



Moreover, not only are the hind coxae large, consisting each of a transverse pale 

 portion, quite concealed within the ventral chitinous wall of the body, as well as of the 

 small globular apical part externally visible ; but the middle and front coxae are of the 

 same form ; so that none of the coxae are really globular, but are all transverse, with 

 only the globular ending visible. This 1 have found to be the case in dissections of 

 Fallia elon^ata and of the Central American Discoloma vestitum Pascoe. The exact 

 condition is described in detail below, under the heading of the genus Fallia. 



Fallia is therefore removed from the Latridiidae, where it was placed by Dr Sharp 

 (op. cit. p. 629), and placed in Discolomidae with Discoloma. 



Notiophygits and Holophygiis were also placed by Dr Sharp (op. cit. pp. 495, 497) 

 in Discolominae (as a sub-family of Colydiidae). In the form of their antennae and feet 

 they agree with Discoloma and Fallia, and it is stated that the affinity of Holophygus 

 with Discolo7)ia cannot be doubted. Altogether it seems probable that the coxae of these 

 genera will prove similar to those of Discoloma and Fallia. 



It is also extremely probable that Aphanocephalus Wollaston belongs to the same 

 family. It was formerly placed in Corylophidae, but Matthews (op. cit. pp. 2, 3 and 

 197) has shown that it differs from them in almost every respect. He erected for it 

 a family Pseudocorylophidae. The affinity of Aphanocephalus to Fallia has already been 

 pointed out (Sharp, op. cit. p. 629). The underside of Aphanocephalus as figured by 

 Matthews (op. cit. PI. VII. fig. C 2) has in respect of structural characters a strong 

 resemblance to that of Fallia : dissection would probably prove the somewhat distant, 

 apparently globular coxae to be in reality transverse. If this should be so, and Aphano- 

 cephalus is placed in the same family as Discoloma, the name Pseudocorylophidae will 

 have to give place to the older term Discolomidae. 



Horn also called attention to the fact that the tarsi of Discoloma and Notiophygus 

 are 3-jointed. Holopliygus, Fallia and Aphanocephalus have also 3-jointed tarsi. All 

 these genera have i -jointed clubs to the antennae. It seems unwise however to mention 

 these facts as characters in the definition of the family, at any rate at present. The 

 matter requires thorough investigation. Aphanocephalus, Holophygus and Notiophygus 

 should be dissected and examined, also a number of other genera. It is possible that 

 some genera may be found to agree with Discoloma in the structure of the coxae, but not 

 in the number of joints in the tarsi and antennae. 



