BEITISH ASSOCIATION* 261 



governed nearly entirely by considerations of number and colour, charac- 

 ters of a very minor importance in classification, though oftentimes con- 

 venient when conjoined with the more permanent characters, of form, 

 position, and structure. Some of his genera probably must fall, at least 

 in their present shape, their descriptions being so imperfect as to render 

 them unrecognisable, owing to his imperfect means of observation: \thus, 

 Trichoniscus may possibly turn out to be Itea, Zia, or some of their allied 

 genera, more fully described by later writers. $ 



Milne-Edwards has for the most part copied Brandt, but has pro- 

 posed new primary divisions, in the naming of which he has been sin- 

 gularly infelicitous. He has drawn his familiar characters from the 

 comparative length or brevity of the last false feet, as compared with 

 the terminal segment of the abdomen, and thus makes two primary 

 groups, Cloportides maritimes (Ligia and Ligidium) : and Cloportides 

 terrestres, subdivided, according to the form and length of the last pairs 

 of false feet and their appendages, intoPorcellioniens (Oniscus, Philoscia, 

 Deto, Porcellio, Trichoniscus, Platyarthus), Armadilliens (Armadilli- 

 dium, Armadillo (includes Cubans), Diploexochus), and Tylosiens 

 (Tylus). 



The names of his primary divisions show us the danger of naming 

 groups from habits. Ligidium, although included among the Maritime 

 woodlice, is as truly a terrestrial species, and found at as great a dis- 

 tance from sea, as Porcellio itself. Most of his descriptions of genera 

 are avowedly mere transcripts from previous authors, even those which 

 one would suppose ought to have come under his own notice, being in- 

 digenous, as in the case of Philoscia, which he allies to Oniscus, whereas 

 in truth it is much more closely allied to Ligia in all its characters. His 

 specific descriptions are almost all mere details of colour, without one 

 truly distinctive character. 



Dana has much extended the group, uniting, however, to the genera 

 already mentioned others, whose claim to the alliance appears but slight, 

 viz., the Limnoriae and Asellidse, on account of the characters of the 

 posterior pair of abdominal appendages. He has thus made one large 

 group, Oniscoidea, and in his definitions has largely entered into cha- 

 racters which may be generic. 



He divides the Oniscoidea into three groups, according to the mode 

 of articulation of the body; the number of articulations of the abdomen ; 

 the size of the last abdominal segment ; the length of the caudal appen- 

 dages (i. e. the last false feet) ; the absence or presence of a palp to the 

 mandibles; and the number of articulations and size of the internal 

 antennae. (Of these, the characters of the number of articulations in 

 the abdomen and internal antennae, the size of last abdominal segment, 

 presence of a palp to the mandibles, have reference solely to his last family 

 (Asellidse), a group which had been probably better omitted.) 1. Ar- 

 madillidae, according to arrangement of caudal appendages, divided into 

 Tylinae (Tylus), and Armadillina3 (Armadillo, Sphaerillo, Armadillidrum, 

 Diploexochus). 2. Oniscidse, according to number and form of articu- 

 lations of maxilliped, form of articulation of fifth joint of external an- 



VOL. iv. 2 M 



