382 OBSERVATIONS AND EXPERIMENTS 



endeavoured to diminish the derivation by the water, by insulating 

 the fish from the water at the moment of the shock. But he 

 attempted this with only the two parts of the fish within the 

 saddle, whereas I proposed the same thing for the part also between 

 the saddles. In this respect the Malapterurus offered great ad- 

 vantages over the long Gymnotus with its lateral movement. 



Fig. 10 shows the contrivance with which I attained my object. 

 It has been not inaptly compared to the cover of a mummy 

 coffin. It consists of a guttapercha cover, which, when the fish 

 rests on the glass plate at the bottom of the tub, fits as closely 

 as possible to its body above and at the side, and touches the plate 



Fig. 10. 



with its lower free edge, a a' is the surface of the water in the 

 tub ; for the sake of clearness, it is shown higher than was usual 

 (see above). In order to prepare such a cover for any particular 

 fish, I cut out a model of the fish in lime wood. This served as a 

 last over which the guttapercha, softened in hot water, could be 

 moulded. When the guttapercha had hardened, it was cut off in a 

 line with the lower edge of the last, the cut edge was again softened 

 and flattened on a glass plate wetted with hot water. A handle of 

 guttapercha is welded on to the back of the cover. Inside this cover, 

 at the ends corresponding to the head and tail of the fish, are linings 

 of tinfoil stuck on with shellac ; these are indicated in the figure by 

 dotted lines. At the ends k and s of the cover, these linings are 

 connected with strips of tinfoil, which, protected by insulating 

 material, pass along the back of the cover to the handle, where 



