CRUSTACEA. 
363 
Any one who has attempted to follow out the development of 
a marine animal from the larval condition is aware how for 
months or years he may seek, and seek in vain, for intermediate 
forms in its chain of growth, although he knows that thousands 
are swimming around him. Surely no one will attribute any 
ver}^ great weight to such an argument. 
He next treats of the subterrestrial Crustacea, and contends 
that none of them can live long out of the water without breath- 
ing air. 
The Decapod Crustacea, which are more or less inhabitants 
of the land, belong to very different families, such as the Ra- 
nidse (Ranina), Eriphidm Grapsoidje [Aratus, Sesarma)^ 
Ocypodidse [Gelasimus, Ocypoda), &c., and must in earlier times, 
according to their capability, have chosen to live in the water 
or not, arguing that those which were air-breathers could not 
have descended from those found to live in water only. 
He then examines the relative persistence and importance of 
the secondary appendage in the Amphipod Crustacea, particu- 
larly in the genus Melita, and argues that its persistence in a 
rudimentary condition in the larva, and sometimes in adult 
animals in which it is assigned to be absent, and its absence in 
certain species that assimilate otherwise in form to those genera 
in which it is present, are arguments strongly in favour of the 
idea that the separate forms come from one common ancestor. 
He draws similar arguments from a new species of Melita, in 
which both pairs of gnathopoda on one side have a small hand of 
ordinary construction, while on the other they are developed into 
enormous pincers*. He contends that all the species of the 
genus Melita^ inclusive of a new one with the large pincers, 
which he has named M. exilii, cannot be the descendants of one 
common ancestor. That is, that all the Melitm with a secondary 
appendage to the antennse descended from a common parent 
tliat is not' the ancestor of M. fresnellii^ and all those which 
had the large pair of gnathopoda descended from a common 
parent that is not the ancestor of M. exilii^ as shown in the 
following diagram : — 
From presence of a secondary 
From form of chelce. appendage to the first antenna. 
M. palmata. M. exilii. M. fresnelii. M. palmata. M. exilii. M, fresnelii. 
III. He next tells us that in thinking out the theory of 
* We suppose that this refers not to both, but to the second pair only. 
