1917.] NATURAL SCIENCES OF PHILADELPHIA. 337 



refraction on plane polarized light, but in an exhaustive treatment 

 of the subject it would be necessary to further consider that plane 

 polarized light, by internal total reflection or reflection from metallic 

 surfaces, at certain angles, may become circularly or elliptically 

 polarized, and in this form pass through the analyzing prism. This 

 may account for the fact that light reflected from the edge of a tin 

 foil cell, used by some preparers in mounting selected diatoms, will 

 appear bright between crossed prisms except where such edge is 

 parallel to the axis of either prism. 



In the communication referred to above, I mentioned that some of 

 the species of Aulacodiscus with the peculiar secondary structure, 

 exhibited bright colors with ordinary transmitted light, which I Avas 

 then unable to account for, but which I would now suggest may be 

 due to optical resonance. The group of phenomena classed under 

 this head, includes the scattering of light by particles or molecules 

 in the atmosphere, causing the blue color of the sky, and similar but 

 not identical effects produced by minute particles embedded in or 

 deposited on the surface of solids. The latter phase of the subject 

 has been extensively studied in connection with the so called ''ultra 

 microscopy" and it is unnecessary to give details here beyond 

 stating that one of the methods of illustrating it consists of dis- 

 tributing minute globules of gold on a glass surface by means of 

 electrical discharges from the end of a gold wire. Such films reflect 

 various colors according to the size of the particles and transmit 

 the complementary color. 



Some years ago, ]Mr. T. C. Palmer called my attention to the- 

 possibility of depositing thin films of gold from a solution of its 

 chloride in collodion, and gave me some of the prepared solution, 

 with which I experimented in depositing the gold on various organic 

 structures with the view of their elucidation, but found the film too 

 coarsely granular to be generally satisfactory. The films, when de- 

 posited on glass, transmitted various colors, generally dark blue, 

 and one in particular, after heating until the surface of the glass was 

 fused, gives brilliant red and blue color effects by transmitted light, 

 reflecting the complementary green and orange tints, due to optical 

 resonance from the minute granules in which the metal was deposited. 

 On examination under the microscope, the granular film shows a 

 startling resemblance to the secondary structure of the group of 

 Aulacodisci previously referred to, and suggests that the color shown 

 by some of them with transmitted light may be due to the same 

 cause. In most of these diatoms the structure is too coarse ta 

 24 



