262 PROCEEDINGS OF SOCIETIES. 



tenure, and form of peduncle and styles of caudal appendages, divided 

 into Oniscinae, Oniscus (including as sub-genera Trichoniscus, Porcellio, 

 Oniscus). Philoscia; Platyarthus; Deto. Scyphacinas (Scyphax, Stylo- 

 niscus). Liginae (Ligia, Ligidium). 3. Asellidae, including Limnorinas 

 (Limnoria). Asellinas (Jasra, Jasridina, Asellus, Janira, Henopomus, 

 Munna). As I believe these last do not really belong to this group, I 

 omit their characters. 



Although this classification contains some new characters, many 

 genera are omitted, which naturally belong to the group, and many of 

 the characters are too artificial to be taken into account. The innova- 

 tion of considering Porcellio, Oniscus, and Trichoniscus as sub-genera 

 is highly objectionable (indeed it will be well when the term sub-genus 

 is banished altogether from our nomenclature), as these divisions do not 

 depend, as he seems to assert, on the mere number of joints in the ter- 

 minal filament of the antenna?, but, as I hope to prove, on a number of 

 other characters sufficient to constitute them, as real genera as Ligia 

 and Ligidum. Mere number plays too great a part in his system also to 

 allow of natural groupings. His classification, however, is valuable as 

 restoring to their proper place the Armadillidas, which, on account of 

 their resemblance to certain Myriapodas, have been hitherto, as it seems, 

 unjustly degraded to the bottom of the list of Cloportides. 



Prom a revision of these various systems, it appears, that putting 

 out of sight such arbitrary characters as mere colour and number, divi- 

 sional characters of most importance are : the comparative development 

 of the external antennas or rather of its filament ; the mode of inser- 

 tion and development of the last pair of abdominal feet, and perhaps 

 we may add the form of the maxilliped ; this last, however, is a cha- 

 racter so difficult of application that it were, perhaps, as well to leave it 

 out. Two very important sets of characters have been omitted or 

 overlooked, which must be taken into account in any classification 

 which seeks to form natural groups ; these are, — the characters drawn 

 from the form of the head; and from the epimerals, as they are generally 

 called. These characters, being permanent in their structure, presenting 

 differences easily caught at, and running also in parallel arrangement 

 throughout the groups, I would propose as the basis of the classification, 

 which, till a better shall arise, I would suggest as that which appears 

 most natural and most in accordance with homological affinities. 



The characters, then, which must be considered in detail are : — 



1st. The amount of development of the so-called " epimerals" ovcoxm 

 of the posterior cephalo-thoracic and abdominal appendages. 



2nd. The form of the head as regards the absence or presence of what 

 are called the frontal or median, and lateral lobes. 



3rd. The relative structure and development of the antennas, and 

 especially of the terminal filament, or, as we will call it, " the tige" of the 

 external or superior antennas. 



4th. The structure and position of the last pair of abdominal false 

 feet, their relations to each other and the last abdominal ring. 



