1901} MICROSCOPICAL JOURNAL. 275 - 
tive of .25 N. A., were bronze-yellow when dry, yellow- 
ish gray in alcohol, blueish gray in medium of 1.41 R.1., | 
iridescent blue in medium of 1.44 R. I., deep greenish 
blue in cedar oil, dark green and pink in. oil of cassia. 
The question of colors shown by diatoms in direct light 
has recently been treated in the Journal of the Quecket 
Club, with special reference to Actinocyclus ralfsii, by 
E. M. Nelson, who has shown that the color cannot be 
due to diffraction. The two valves of A. ralfsiz which 
were included in the previously described slide showed 
only pale brown and grayish tints in media of R. I. be- 
low 150, and extinguished with the other forms in one of 
R. I. about 1.43. In cedar oil one valve showed a blue © 
color and in oil of cassia both became brilliant with green, 
blue, purple and yellow. Under wide aperture objectives 
the color is not visible when diatom is sharply in focus, 
but it appears as soon as thrown slightly out of focus. 
This color appears to be due to dispersion, and its nature 
and cause might possibly be further elucidated by study- 
ing the effect produced by different media such as were 
employed in this case.—Proc. Phila. Acad. Natural, 
Sciences. | 
Diatoms, The Food of Fish in Kansas. 
GEO. H, CURTIS. f 
Mr. 8. G. Mead, of McPherson, gave me a small fish 
about two inches long, which he caught at Belvidere, 
Kiowa county, Kansas, last fall. It was apparently a 
young perch, to judge from its shape and the dark bands 
along its sides. Having a curiosity to know what its 
food had consisted of, I undertook a microscopical examin- 
ation of the contents of the digestive tract; but the dif- 
ficulty of arriving at satisfactory results was much in- 
creased by the carbolic acid and oil the fish had been pre- 
served in, which interfered very much with the proper 
