328 Natural History Bulletin. 



The foot is of smaller size than that of D. marginalis and 

 offers another example of the correlation between size of 

 patellas and number of palettes spoken of in the remarks 

 under Cybister tri^unctatiis. This parallel does not however, 

 hold good in such forms as D. harrisii and others where the 

 small palettes are comparatively much less developed, and 

 will probably apply only between closely allied species. 



Dytiscus circumcinctus Ahr. Plate VII, Fig. 5. 



Patella nearly circular in outline, not elongate. The pal- 

 ettes are smaller and more numerous than in either of the 

 preceding species, the marginal row of the first joint contain- 

 ing twenty-one, while the angle bears thirty-five others. The 

 second joint bears five illy defined series of them instead of 

 the two seen in D. fasciventris and D, margmalis, while the 

 third joint has a great number placed confusedly with scarcely 

 a tendency to arrangement in lines. 



Dytiscus punctulatus Fabr. 



Resembles D. circuincinctus, but the small palettes are still 

 more numerous, especially on the second joint where about 

 eight much confused rows can be made out. It approaches 

 the next species in structure. 



Studied from European specimens. 



Dytiscus harrisii Kirby. Plate VI, Fig. 3. Plate VIII., 



Fig. 9. 



Patella slightly elongate, third joint more deeply notched 

 at apex than in the other species. The small palettes are 

 minute, very closely placed and extremely numerous. There 

 are two nearly complete rows on the basal joint, the second 

 has about ten, very much confused, for the greater part of 

 the width. 



This species shows a slight indication of a bare median 

 space on the third joint, reminding us of the structure shown in 

 the foot of Acilius. The middle tarsus as figured shows very 



