( 12 ) 



are inserted internal to, and at the base of, the external antennae ; they 

 afford but little assistance in classification, for although they have 

 been made use of to distinguish genera, in the higher divisions they 

 are useless as diagnostic characters. We pass then to the next 

 or — 



External antenna. — These are invariably simple, being made up of 

 a peduncle, which is pretty constant in its characters, and a filament or 

 tige, which varies much, and therefore affords good classificatoiy cha- 

 racters. The peduncle invariably, as far as we know, consists of five 

 joints: in some, as Ligia, a sixth appears to exist, but a little examina- 

 tion will show this to be a part of the third ring. The only joint which 

 affords any useful character is the second, which, in some genera, as 

 Armadillium, Porcellio, and Oniscus, is narrow at its articulation with 

 the first ring, and then suddenly swells out into a broad expanse, con- 

 tracting again slightly towards its articulation with the third joint; 

 while in others, as Philoscia, Ligia, Philougria, Itea, and Ligidium, 

 the second joint is globular : another argument for the separation of 

 Philoscia and Oniscus. The carvings and sulcationson the joints of the 

 peduncle, however, occasionally afford useful specific characters. 



From the extremity of the peduncle arises the filament, and of this, 

 trusting to form alone, the seven following types have been described. 

 This organ has attracted much attention ; but unfortunately, as I said 

 before, too much attention has been paid to mere number, and too little 

 to form. 



Regarding form only, we find the filament constructed on three 

 types:— 



1st. The articulations are short, globular, cupped at their upper ex- 

 tremities, which are fringed with hairs, and receive the inferior portion of 

 the articulation next succeeding. These filaments are all, as far as I 

 can learn, multi-articulate. Examples — Ligia, Ligidium, Titanethes, 

 Stjdoniscus. 



2nd. The articulations elongated, somewhat flattened, and gene- 

 rally covered with hair, few in number, gradually tapering, but not 

 subulate, the terminal one terminating in a narrow articulated hair. 

 Examples — Oniscus, Porcellio, Philoscia (?), Armadillium, Armadillo, 

 Spherillo, Tylus, Scyphax, Platyarthus (?), Deto (?). 



The characters of the second articulation of the peduncle, as hinted 

 above, divide these into two groups. 



3rd. The articulations few in number, the whole filament gradu- 

 ally tapering, subulate, and perfectly naked, terminating in a tapering 

 filament. Examples — Philougria, Itea, Trichoniscus (?). 



In all these genera the number of joints in the antennae has been 

 used as a generic character by Brandt, &c., except Ligia, Ligidium, and 

 Platyarthus ; but in grouping the genera no regard has been, as far as I 

 know, paid to the far more important character — that of the form of 

 the tige. 



Dana states that number of articulations of the tige is of no value as 

 a character, and instances as a proof it the genus Scyphax, in which he 



