424 FRANCIS B. SUMNER 



regards the fourth of these specimens (no. 2), it will be seen by a 

 comparison of figs. 2e and 2e that the appearance upon the coarse 

 gravel was somewhat different from that upon the fine, and that 

 the difference, so far as it went, seemed to be in the direction of a 

 greater adaptation to the former. The case is not, however, a 

 particularly striking one, since we should regard the appearance of 

 the fish in fig. 2c as tolerably well adapted to the finer material. It 

 is interesting that the coarsely blotched effect is here produced in a 

 somewhat different manner from that shown in fig. le, since, in 

 the present case, the pale spots are clearly in view all over the sur- 

 face, and form a conspicuous part of the skin pattern. 



Certain other materials were employed as backgrounds, which, 

 although 'natural,' in the sense of occurring in nature, were not 

 found in any part of the possible habitat of the species under f 

 consideration. ""* As the most important of these, we have : 



4. A very fine, jet-black sand, composed almost wholly of 

 crystals of magnetite. This was found in the bed of a stream in 

 the island of Ischia,and had to be separated bymeans of consider- 

 able washing from the paler sand and mud with which it was mixed. 

 Although of nearly a jet-black shade, these crystals gleam so 

 much in certain lights that the sand may appear to the observer 

 to be far from black (fig. Ic). My chief object in using this mate- 

 rial was to test the question whether on a bottom devoid of white 

 particles^-* the white specks forming the annuli would completely 

 disappear. Three specimens were put to this test. One of these 

 was the individual which I have designated as no. 1, this being 

 a specimen which had yielded some of my most striking results. 

 Fig. Ic shows the condition of the fish after a stay of four days-^ 



2' This is particularly true of the magnetite sand. On account of the high specific 

 gravity of this substance, any other ingredients with which it was mixed,— and these 

 were all much lighter in color — tended to separate out and come to the surface. 

 It thus seems improbable that an uncovered bed of such material should be formed 

 anywhere upon the sea-bottom, and could ever constitute a part of the normal en- 

 vironment of any fish. 



2* This could of course be likewise tested by placing the fish on a black painted 

 bottom. But the consistency of the sand seemed better adapted to calling forth 

 normal reactions. 



^* According to my notes, the maximum effect was attained after two days. 



