426 FRANCIS B. SUMNEB 



even during a stay of eight days upon this material. Another 

 specimen (fig. 6c) assumed much of the color as well as the texture 

 of the sand. This fish which, my notes state, had exhibited very 

 little red in its composition when kept on other backgrounds 

 assumed, after six days on the sand, a general tone resembling 

 the no. 122 of Klincksieck and Valette. Fig. 6c does scant justice 

 to the extent of the harmony. A third fish (no. 4) was kept upon 

 this sand for a period of six days. Although, in general, this 

 specimen had been a highly adaptable one, it proved to be entirely 

 refractory upon this material, changing little, if any, from its 

 previous condition, which was quite ill-adapted to the present 

 type of bottom. 



A very coarse gravel, likewise containing considerable red in' 

 its composition, was employed, but only one or two specimens were 

 tested upon this, and the results w^ere negative. 



6. A white sand-like preparation obtained by grinding up 

 fragments of marble. Upon this, certain specimens became very 

 much paler, though none of the three fishes tested attained the 

 same degree of pallor as those which were kept upon the marble 

 bottom of a large aquarium or upon white glass plates. Indeed, 

 one of the three did not undergo any striking change, even after 

 11 days. 



7. In another experiment, large, rounded fragments of nearly 

 black lava were used in combination with this white artificial 

 sand, giving a vividly contrasted effect. The result, in the case of 

 specimen no. 1 (the only one tried) was the appearance of large 

 dark blotches and a coarsely mottled effect, resembling that shown 

 in figure le or figure 1/. 



8. A background just the reverse of the last was obtained by 

 the use of rounded fragments of white marble imbedded in the 

 coarse dark sand (1) or in the magnetite sand. The only experi- 

 ment in which the latter combination was employed gave negative 

 results, as has already been stated. The marble fragments and 

 coarse dark sand were used with specimen no. 6, and the effect 

 is shown in figure 66. A considerable measure of contrast was 

 brought about in the skin of the fish, but on the whole less than I 

 had expected. The pale spots were at no time even approximately 

 white. 



