ART. 23. BLISTER BEETLE TRICRANIA PARKER AND BOVING. 17 



tarsal tip as proximally from tarsal base. Tarsus moved by a flexor 

 and a retractor, (m^ and m^, fig. 6) but no muscles attached to the two 

 setae at the base.* 



First to seventh abdominal segments, with tergal shields sub- 

 trapezoidal: first abdominal segment about four times wider than 

 Jong; segments gradually decreasing posteriorily in length and 

 width ; seventh abdominal segment about half as long and not much 

 more than one-fourth as wide as the first one ; latero-posterior angles 

 of tergal shield in all segments alm.ost rectangular, carrying a short 

 terminal seta. 



Eighth abdominal segment, subquadrate. From posterior margin 

 of the tergal shield one pair of falciform, distally attenuate, flat 

 processes extends backward about as far as the segment is long; they 

 are diverging, their convexities opposed, at base as far apart as half 

 their length and distally twice as far or more; each is adjacent to the 

 inner and upper side of an equally long but somewhat wider and 

 conical hook, which projects slightly below the falciform process. 

 Apically and facing the process each hook carries the annular peri- 

 treme of the eighth abdominal spiracle, and inwardly contains the 

 elongate, ovoid spiracular atrium (fig. 2). Undoubtedly the object of 

 these structures is to facilitate the breathing under extraordinary 

 and difficult conditions. The function of the adjacent processes may 

 possibly be that of holding a supply of air between themselves and 

 the ends of the hooks to prevent water from penetrating into the 

 tracheal system at the period when the larva is free living in the field 

 (see p. 5). Later on, when the larva has invaded the cell of 

 CoIIetes, the air supply nnij prevent honey from obstructing and 

 clogging the spiracular opening, if accidentally the larva should be- 

 come temporarily submerged.^ 



Ninth abdominal segment, cylindrical, two-thirds as wide, twice as 

 long as the eighth abdominal segment ; terminally on each side, with 

 one rather short macrochaeta of same length as the segment. 



* These setae in the Meloid larvae have been interpreted and termed as " claws " by 

 most authors ; but erroneously, because true claws, as developed for instance in Cara- 

 bid larvae, are provided with individual muscles. 



^ The interpretation of the object of the hooks and processes, given hy Fabre, Mayet, 

 and Beauregard, who consider them as locomotory and grasping organs, has already been 

 abandoned by Cros in several of bis masterly papers, for instance on page 59, iu his 

 study on Ncmognatha cliri/someUna (Bull. Soc. Hist. Nat. Afrique du Nord, vol. 10. 1919). 

 Cros calls the organs " appareil respiratoire erectile dorsale " and probably looks upon 

 their physiology and use as here described. He has, however, avoided proposing any 

 definite theory on this subject, unless such is found in his study on Hornia nymplwides 

 (Bull. Soc. Hist. Nat. Afrique du Nord, vol. 5, 1913), which Is not available to the 

 present writers. To this study Cros refers in his " Larves primaires des Mcloidae," (Ann. 

 Soc. Ent. France, 1919, p. 262) as follows : " In my paper on Hornia nymphoides Escal., 

 I have discussed at groat length the structure and the functions of these organs to 

 which J. H. Fabre has assigned, but by no means proven, the rdle of fixation and loco- 

 motion ; in reality t'-iey are but mndiiied spiracles." 



When Cros mentions the organs as " erectile," the present writers must admit that they 

 never have been able to see them make actual motions. 



