IRIDESCENT COLOES OP BIRDS AND INSECTS MALLOCK. 427 



low intensity is of course clue to the small area of each convex surface 

 which reflects light in any given direction. 



In attempting to investigate the origin of the colors many methods 

 were employed, the first and most obvious being to cut thin sections 

 normal to the color-producing surface and then to examine them 

 with the highest microscopic power available. If the colors are 

 analogous to those of thin plates, it is -clear from the high intensity 

 of the reflected light that more than one pair of surfaces must cooper- 

 ate in the reflection. In general the reflected light is not even 

 approximately monochromatic, and this fact limits the number 

 of surfaces which can be supposed to act, but if the surfaces are 

 supposed to be separated by air and placed at the most favorable 

 intervals their number need not exceed three or four to account 

 for the observed intensity and tints. 



The most favorable spacing for the successive layers is that their 

 thickness and the intervals between them should be a multiple of the 

 half wave-length of the mean ray, reckoned in the length of the waves 

 within the material of the layer, and it was thought possible that the 

 thin sections might show a laminated structure. 



For the material of feathers and insects' scales, n is somewhere 

 about 1.5 or 1.6, so that the least thickness for the plates of refractive 

 material would be of the order of one one-hundi'ed-and-fifty-thou- 

 sandth and the air intervals one one-hundred-thousandth of an inch — 

 both beyond the resolving power of the microscope; but from the 

 composition of the reflected light it seemed likely that the intervals 

 might be two or three half wave-lengths, which would be readily seen 

 as far as adequate separation of the images is concerned. In nearly 

 all the sections examined bands of this order of thickness appeared 

 with some forms of illumination, but it was impossible to be sure that 

 they were not due to diffraction effects from parts of the section slightly 

 out of focus. 



There are many difficulties in preparing sections thin enough for the 

 advantageous use of objectives with large angular aperture. When a 

 section is to show a stratified structure its thickness should certainly 

 not be greater than the distance between the successive strata, and 

 may with advantage be much less. It was not difficult to cut sections 

 about one twenty- thousandth of an inch thick, but this is three or 

 four times too thick to show with certainty stratification whose pitch 

 is one sixty- thousandth or less. 



Occasionally, by accident, thinner sections (perhaps one forty- 

 thousandth) would be cut, and these showed apparent stratification 

 most plainly, but in no case was the image free from the effect caused 

 by some part of the thickness of the section being out of focus, and, in 

 all Drobability, what appeared to be stratification was in reality a series 



