160 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1911. 
its lowest reflecting power and greatest transparency has been so 
tremendously weakened by atmospheric absorption, that it is neces- 
sary to employ thick films and long exposures, otherwise the action 
upon the photographic plate results chiefly from the violet and ultra- 
violet rays, which are capable of traversing glass. 
As an illustration of the behavior of silver films of different thick- 
nesses, used as ray filters, we may take some pictures which were 
made for the purpose of studying the reflecting power of various 
metals, suitable for telescope mirrors, for ultra-violet photography. 
As silver reflects only about 4 per cent of the ultra-violet in the 
spectrum range for which it is transparent, a silvered glass reflecting 
telescope for this purpose is obviously out of the question. Speculum 
metal is fairly suitable, but speculum mirrors of large size are trouble- 
some, and difficult to procure. I accordingly worked out a method 
of depositing nickel on glass. The glass is first silvered, and then 
electro-plated with nickel, by a process which I have described 
recently in the Astrophysical Journal (Dec., 1911). The double 
sulphate of nickel and ammonia is used with one or two dry cells. 
The solution must be very dilute (10 grams or less to the liter), 
otherwise the nickel strips the silver from the glass. We have here ~ 
four pictures of a silvered glass dish, partially plated with nickel 
(pl. 8, fig. 1). The silvered portion is marked Ag, the nickel Ni, 
while at G we have a spot of clear glass from which the metal has 
been removed. The dish stands against a flat plate of polished 
speculum metal Sp, and the metal surfaces reflect the light of the 
sky to the camera. The first picture was made by blue and violet 
light without any ray filter, and as you see the glass surface G is 
quite black, while the silver reflects much more powerfully than the 
nickel. The following three pictures were made with a quartz lens, 
coated with silver films of increasing thickness. The silver and 
nickel reflect to about the same degree in the second picture, in the 
third the silver is much darker than the nickel, while in the fourth 
the silver is seen to reflect no more than the spot of clear glass G. 
This last was made through a film, through which a tungsten lamp 
was invisible. If these ultra-violet rays were visible to us, metallic 
silver would appear to have about the same reflecting power and 
appearance as anthracite coal. 
We will next take up the action of our atmosphere on these ultra- 
violet rays. J have taken two photographs of a man standing in the 
road in full sunshine, in the one case by ordinary light and in the 
other by ultra-violet radiation. In the latter the shadow is com- 
pletely absent. Ultra-violet behaves in exactly the opposite way to 
the infra-red. The infra-red rays are enabled to drive through the 
atmosphere without being scattered laterally by the molecules of the 
air or the dust particles. The short or ultra-violet rays, on the other 
