146 =Mr.J.S. Baly on the Classification of the Eumolpide. 
melide which I collected numbered only 55 species, belonging to 
4 genera; whereas the Eumolpide reached the large figure of 250 
species, comprising a large number of genera. In Europe this pro- 
portion is more than reversed ; for the Chrysomelide number 236 
species (11 genera), whilst the Eumolpide are represented only by 18 
species (7 genera). The causes of this difference may perhaps lie in 
the circumstance that the Chrysomelide feed principally on shrubs, 
which form a large proportion of European vegetation, but constitute 
a subordinate feature in the equatorial forests, whilst the Eumolpide 
live on trees, whose proportion to the shrubby vegetation is immea- 
surably greater in the tropical forests than in Europe. Little can 
be said regarding the habits of the Eumolpide, all the species being 
very similar in their modes of life. A large number of them, how- 
ever, seem confined to certain trees ; and it is possible that consider- 
able diversity might be found, were the food plants of each kind 
well ascertained, and their development from the egg to the perfect 
state carefully traced. Many of the larger metallic species are found 
only on arborescent Solanacee, or plants of the Potato order, which 
grow in all waste places and neglected gardens in the suburbs of 
towns and villages. All are gregarious in their habits, like the true 
Chrysomelide ; and although they are not so obese in habit and slow 
in movement as those insects, they seem to make quite as little use 
as these of their power of locomotion. As they do not generally 
feign death and drop to the ground on the approach of danger, like 
the Clythride and many of the Chrysomelide, and have not the strange 
disguises of the Chlamyde, or the tenacity of grasp on foliage of 
the Cassididee and Hispidee, or the nimble flight of the Megalopide, 
the reason of their existence in such large numbers in situations ex- 
posed to the depredations of birds and lizards may perhaps lie in 
their having some passive means of defence of which we are at pre- 
sent ignorant.” 
I propose breaking up the family into a number of subfamilies, 
founded rather on natural affinity than on technical characters. 
Division I. Anterior episterna variable in shape, always prolonged 
backwards along the outer edge of the anterior coxe. 
Subfamily I. Apoxrn 2. 
Body oblong or elongate, subcylindrical, non-metallic, clothed above with 
hairs or scales. Thorax commonly subcylindrical ; its lateral margin 
usually obsolete, more rarely indicated by a faint ridge or irregular 
teeth ; its lateral surface generally forming the segment of a circle with 
