224 Mr. R. I. Pocock on a new Genus 



keels testaceous, lateral portions of the segments darker but 

 mottled, the darker shades occurring in patches. Most of the 

 large dorsal tubercles almost black, some bright yellow, a few 

 dull-coloured ; the yellow tubercles in patches of two, three, 

 or four together, but not occurring at definite intervals. 



Two female specimens from Perak in the Malay Penin- 

 sula. 



In 1872 M. de Saussure, basing his classification upon the 

 form of the head and jaws, divided the family Polyzonidas 

 into two tribes- — the Platydesmia, to contain Platydesmus, 

 and the Poljzonia, to contain Polyzonium and Siphonophora ; 

 and Dr. Latzel in 1884 also divided the PolyzonidgB into two 

 subfamilies, for one of which he adopted the term Platydesmia, 

 while to the other he gave the name Dolistenia, But the 

 latter author, considering the number of body-segments to be 

 a character of systematic value, included in the Platydesmia 

 all those suctorial Myriopods which possess fewer than seventy 

 segments, while the Dolistenia contained all those forms in 

 which the body is composed of more than seventy segments. 

 This arrangement brought about the association of Polyzonium 

 with Platydesmus and the separation of Polyzonium from 

 Si^]w7iop]iora, and if adopted in the present case would lead 

 me to assign to Pseudodesmus a place, not with Platydesmus^ 

 but with Biidliono'phora^ thus showing that, in my opinion, 

 the relationship between Siphonophora and Pseudodesmus is 

 greater than the relationship between the latter and Platy- 

 desmus. But that is not the case ; the form of the gnatho- 

 chilarium shows that Platydesmus and Pseudodesmus are 

 closely allied, and the form of the proboscis shows that Poly- 

 zo7iium and Siphonophora are closely allied. These two 

 things, and the knowledge of the fact that the number of seg- 

 ments, being very variable, is a character practically valueless 

 for classification, have led me to reject the divisions of Dr. 

 Latzel and to adopt, at all events provisionally, the older ones 

 of M. de Saussure ; but at the same time it seems to be very 

 probable that careful examination of the mouth-parts of genera 

 that have hitherto been but poorly described will, by bringing 

 to light intermediate forms, render impossible the attempt to 

 divide the Polyzonidas into groups larger than genera. 



Owing to the scanty descriptions which at present exist of 

 the following forms — Octoglena (Wood), Petaserpes (Cope), 

 Andrognathus (Cope), and Dolistenus (Fanzago) — it is impos- 

 sible to associate them with either of the subfamilies adopted ; 

 and it must be confessed that the classification of Dr. Latzel 

 highly commends itself from the fact that in formulating it 

 the author was able, inasmuch as the number of segments 



