285 JOURNAL, BOMBAY NATURAL HISTORY SOCIETY, Vol. VIit. 
simple and available in its application, yields unvarying results with different 
observers—results, moreover, which admit of the simplest description. 
Before describing this method, I may say that long experience in the 
examination of iridescent objects has proved to me that, almost without 
exception, the colours of natural iridescent objects are due to interference 
produced by thin plates. In order, therefore, to render clear the principles 
on which the method I propose is founded, I will briefly refer to certain 
fundamental facts in connection with colour production by thin plates, and 
for this purpose will select a thin film of mica, which, with light at 
perpendicular incidence, appears red, iridescent red. If, now, this plate be 
inclined so that the light falls on it at a more oblique angle, it is, of course, 
reflected atthe same angle, and now appears orange, and if the plate be still 
further inclined, the reflected light appears yellow, then yellowish-green, 
green, and bluish-green ; and if the light were not too copiously reflected from 
the first surface to allow of perceptible interference by further inclination of 
the plate, all the colours of the spectrum in their proper sequence might be 
observed. The same results, but much more vividly, may be seen in these 
erystals of chlorate of potash. Thus we see that by rendering the incident 
light more and more oblique, the reflected light changes from a lower toa 
higher tint, that is, from the red towards the violet end of the spectrum, And 
this is what occurs in the case of all iridescent bodies ; as the incident light 
becomes more oblique, the colour changes to the tint above it in the spectral 
order, so that, if we know what colour any such object appears when seen ata 
certain angle, we can infer what colour it will change to on varying the 
incidence. This beetle (Sagra purpurea), for instance, is red at perpendicular 
incidence ; it will, therefore, appear orange-yellow and green when examined 
by successively increased obliquity of light. And the same is true of all other 
iridescent red objects. If the object at perpendicular incidence be green, as 
in the case of this beetle (Buprestis), it will become blue and then violet as the 
incidence is increased. We thus see that an iridescent object varies in colour 
simply because it is examined by light incident, and therefore reflected, at 
different angles. Thus, different observers see the same iridescent object of a 
different colour, when they view it illuminated by light at a different angle of 
incidence. If, however, the object is seen by all at the same angle of the 
incident light, it will present the same colour ; and this is, in fact, what the 
method I propose ensures, #.¢., that iridescent objects shall always be seen by 
light at one and the same angle of incidence. The angle I select is one of 90°, 
so that the incidence and reflection are normal or perpendicular to the reflect- 
ing surface. By selecting this angle, all trouble of measuring angles is 
avoided, since we know that the incidence is perpendicular when it coincides 
with reflection. Now the reflected light may be made to coincide with the 
incident light by reflecting it on to the object by means of a mirror, and so 
adjusting the object that the light reflected from it passes to the eye through 
