Xll INTRODUOTOKY REMARKS. 



Geodephaga, on the contrary, is, in proportion to the area, rather 

 less so. The following Table, as compared with the corresponding 

 one in the Madeiran Catalogue, will show this more clearly : — 



HJn/iicJwphoi-a 17(5 



Brachchjtra 141 



Heteromera 125 



Necrophaga 114 



Gcodephaga ll-> 



Priocerata 89 



Cordylocerata 51 



Phijtopliaga 44 



Hydraclephaga 22 



Philhi/dnda 21 



Pseiulotrimera 19 



Eucerata 15 



930 



Of the genera, as yet detected at the Canaries, the largest (and by 

 far the most characteristic) is Lajicirocerus — of which no less than 

 35 exponents have already been brought to light (and there ai-e pro- 

 bably many yet to be found)*. The next is Homalota ; but as a 

 considerable proportion of the minute Staphylinids which compose 

 that immense group are erninently liable to accidental diffusion 

 (through indirect human agencies) over the civilized world, I lay but 

 little stress upon this fact. But the third in order, namely Hegeter, 

 is quite as significant as the first — numbering no less than 19 species f. 

 Then follow Calatlius and Attalus — both of which are largely ex- 

 pressed, and have 17 representatives (manifestly quite indigenous). 

 Apion has 15, but some of them may perhaps have been introduced. 

 Bemhid'uim, Artlirodes, and PMlonthus have each 14; Accdies and 

 Longitarsus 13 (the former being equally developed, or even more so, 

 at Madeira) : Sa2ynmis, Phnelia, Helops, and Anthicus number, each 

 of them, 12; Aphanarthrum W (all («7</'rt-indigenous) ; Hgdroponis 

 10 ; and Tarpliius (likewise positively endemic, and of which more 

 will doubtless yet be found) 9. Indeed Tarphius, as I have elsewhere 

 shown, is almost characteristic of the intermediate sylvan districts of 

 the whole of these Atlantic Islands ; nevertheless it is decidedly more 

 dominant at Madeira (where no less than 20 exponents have already 

 been observed) than at the Canaries. 



* I include with Laparoccrus my genus Athinfis. 



t I include ThalpophUa and Gnophota (which arc scarcely more than sub- 

 genera) with Hegctcr. 



