12 
October, 1852 (see also Zool. 3633), my friend, Mr. P. H. Gosse, 
who was then sojourning at Ilfracombe for his health, offered in the 
kindest manner possible to supply me with materials, and from that 
period he has always most heartily responded to my wants. It must 
not be imagined for a moment that the beautiful creatures I have thus 
received have been all preserved alive or always quite healthy. In 
experimental investigations this would be unreasonable to expect, as 
the very fact of experimenting implies a disturbance of the then state 
of things. Besides which, from want of a sufficient knowledge of 
Natural History, from want of forethought and experience and other 
causes, I have lost many very fine specimens; and as the detail 
of these losses may prevent the occurrence of the like annoyances to 
others, I shall venture to occupy your time for a short period with 
their history. 
My greatest loss arose from too great an anxiety to transfer the col- 
lection I had preserved in a healthy condition to the end of December, 
1852, into the new tank. As soon as it arrived from the maker's I lost 
no time in introducing my numerous family to their new abode, and 
dearly I paid for my precipitancy, for on the next morning I found 
many of my most beautiful specimens dead; thus I lost two fine 
Holothurias (HZ. Pentactes), a small freckled goby (Gobius minutus), 
a beautiful little pipe-fish (Syngnathus lumbriciformis), and several 
others, and on opening the door of the case the cause of this mor- 
tality was at once evident,—an iridescent film of oily matter was float- 
ing on the surface of the water, arising from the paint with which the 
angular joints and edges of the small tank had been coloured not 
having become sufficiently hardened. 
Another source of loss arises from the several creatures attacking 
and devouring each other, and it therefore becomes a point of great 
importance—and highly necessary to be carefully observed, where 
their preservation is an object—to ascertain what varieties may 
be safely associated in the same tank ; as, for instance, I have found 
that the shrimps and prawns attack, and very soon devour, all 
the larger varieties of corallines and Polyps, Sabelle, Serpule, rock- 
borers, Cirrhipeds, some of the Annelids, many bivalve and univalve 
mollusks that are unprotected by an operculum, or have no power of 
closing their valves. The instances which have come under my own 
immediate observation have been the destruction of the Pholas dacty- 
lus, Saxicava rugosa, Cyprea Europa, and several specimens of 
Sabelle, Serpule, Coryne sessilis, and many others. 
The common crab (Cancer Mfenas) is likewise a most destructive 
