THE PHOTOGRAPHY OF FISH. 63 



firstly, that an enormous amount of time and patience 

 must have been expended on the illustrations (Mr. Day 

 tells us in his preface that he " personally delineated 

 every species from nature") ; and, secondly, that by a. 

 natural consequence the completion of each plate must, 

 in most cases, have been long time subsequent to the 

 death of the fish. The change in shape and appearance 

 of our common fish, which follows immediately on their 

 removal from the water, cannot have escaped the notice 

 of the most casual observer. As a consequence of this 

 change, the correctness of a fish drawing must, so far 

 as contour is concerned, depend on the correctness 

 either of the artist's memory, or of the description with 

 which he is supplied. In cases where the artist has 

 not the advantage of even a moderately fresh specimen, 

 but has to depend on one which has been stuffed or pre- 

 served in spirits, the resultant drawing becomes almost 

 entirely a matter of guesswork. It has been well pointed 

 out that, in addition to the inherent want of accuracy in 

 such a class of illustration, the expense of original hand- 

 work leads to a constant repetition of error, as one fresh 

 natural history succeeds another with borrowed illustra- 

 tions. An absolutely true representation of a fish can 

 only be obtained by a photograph of it while alive, and 

 as far as practicable under normal, that is, swimming, 

 conditions. If it be impracticable to obtain the live 

 fish, the dead specimen should be photographed with 

 as little delay as possible, so that the representation of 

 the general contour may approximate to reality. A 

 long period of " stuffedness," or even the delay of a 



