OF MARINE AQUARLE, 357 



they are apt to get out of order, unless made by skil- 

 ful and thoroughly competent artistes. Under 

 certain circumstances, there is no doubt that an 

 oblong tank of moderate dimensions is a great 

 desideratum ; but what I wish particularly to 

 impress upon the mind of the reader is, that a large 

 tank is not at all necessary in order to study the 

 habits of marine animals; indeed, the more capa- 

 cious the vessel, the more difficult becomes the task 

 of watching the secret movements of any of its occu- 

 pants. On this account it not unfrequently happens 

 that a common glass tumbler becomes of much 

 greater service to the student than the most elabo- 

 rate aquarium. 



The tanks which I use are circular in form, the 

 largest being not more than sixteen inches in 

 diameter, by seven inches in depth. Its cost was 

 four shillings. Each one rests on a base of mahogany, 

 elevated on turned legs to a height of nine inches. 



Some persons object to the circular tank, on the 

 ground that its occupants when seen from the sides 

 appear magnified. This fact, as I have elsewhere 

 remarked, is rather a recommendation with me, as 

 it presents more distinct views of each movement 

 in the vessel, and whenever I wish to see the objects 

 of their natural size, I can do so by looking in from 

 the top. 



On the edge of the tank are placed three chips 

 of gutta percha, in which are inserted three steel 



