1911.] NATURAL SCIENCES OF PHILADELPHIA. 113 



light. In the absence of such black opals, a good substitute can be 

 prepared by grinding a thin section of ordinary precious opal and 

 backing it with India ink. Some opals in a dark matrix, likewise serve 

 admirably. 



On examining the colored reflections with a micro-spectroscope, 

 instead of a spectrum with a dark band in the position of the com- 

 plimentary color, which characterizes the interference spectrum from 

 a thin film, we find only a narrow band of same color as that shown 

 visually. This has been explained by Lord Rayleigh as being due to 

 the successive action of numerous parallel films of a thickness of the 

 same order as the wave lengths of light, an explanation first applied 

 to the similar colors reflected by certain crystals of potassium chloi ate, 

 in which the films are known to be due to repeated twinning. To 

 what they are due in opal is not known, but under favorable conditions 

 they may become quite visible, although I have hitherto been unable 

 to distinguish them in sections at right angles to the laminae. In one 

 specimen of opal reflecting a brilliant green color have been able to 

 count laminae 38,000 to the inch, and as they appeared to be at an angle 

 of about 45° to the plane of the section this would correspond to over 

 50,000 to the inch at right angles to the plane of lamination. 



The manner in which the colors of opal have been accounted for 

 justifies the application of same explanation to the iridescent feathers 

 previously referred to, which show bright-band spectra, although I 

 have seen none in which the band was as sharp and narrow as in the 

 opal, whose colors are probably the purest shown by any natural 

 object. It will not be safe, however, to apply the same reasoning to 

 all feathers, for Nature does not hesitate to use various means to the 

 same end. An illustration of this will be found in another humming- 

 bird feather, which in general structure corresponds to that already 

 mentioned, except that the reflections are blue in color, but at the end 

 of each pinnule there is a single filament which glows with a most 

 brilliant ruby color and shows a spectrum not unlike that of the other 

 ruby-tinted feather, but on examining a balsam-mounted specimen 

 with transmitted light, the color and spectrum are the same as by 

 reflected light, proving that it is due to absorption and not to inter- 

 ference. The brilliancy will be accounted for by examining sections 

 of the filaments, which prove to be somewhat rounded trigonal prisms, 

 with an edge facing outward, so that light striking either side is re- 

 turned by internal reflection from the back ; or, in other words, Nature 

 has here employed the same method that a skilful lapidary would 



8 



