810 



SCIENCE 



[N. S. Vol. XLI. No, 1066 



tion of photography would bring greater 

 accuracy in stellar photometry, but at pres- 

 ent the errors of the best photographic 

 measures and the best visual ones are about 

 the same. 



The use of the property of selenium as 

 a basis of some form of photometer has been 

 made by various investigators, but not 

 many have tried it on faint objects. The 

 principle of converting a light effect into 

 an electrical one is quite simple, for what 

 we call a selenium cell is a bridge or re- 

 sistance. Light from a bright source like 

 the sun, falling upon a selenium element of 

 1,000,000 ohms, will reduce the resistance 

 to say 20,000 ohms or one fiftieth of the 

 original. For faint lights there are some 

 special electrical connections which give 

 the best arrangement, but let it suffice to 

 state that as used with our telescope a sele- 

 nium cell is connected as one arm of a 

 Wheatstone bridge, that we use a d'Arson- 

 val galvanometer, and current is supplied 

 by a few dry cells. 



The nature of the problem becomes ap- 

 parent when we state that the image of a 

 second magnitude star, say the Pole Star, 

 near the focus of a 12-ineh telescope objec- 

 tive gives the same surface illumination on 

 a selenium cell that would come from a 

 candle at 150 meters' distance, without any 

 intervening lens. Therefore, to measure 

 the light of such a star with a probable 

 error of 1 per cent, is equivalent to the 

 detection of a candle at 1,500 meters, or 

 roughly a mile. In theory, to work with 

 faint lights we might increase the voltage 

 and use a very sensitive galvanometer, but 

 unfortunately selenium is not so uniform in 

 its action that the sensitiveness of an ap- 

 paratus can be increased without limit, and 

 the peculiar irregularities of behavior have 

 prevented selenium from being of wider 

 application. It is especially susceptible to 

 temperature changes, and after exposure to 



light requires considerable time for recov- 

 ery. It becomes more sensitive and extra- 

 ordinarily more regular with decrease of 

 temperature, and conditions are probably 

 best when a cell can be maintained at a 

 uniform temperature of about — 20 degrees 

 Centigrade. 



We have found it best to keep an ice 

 pack about the cell at the end of the tele- 

 scope for work in moderate or warm 

 weather, and the whole apparatus is 

 wrapped up in a blanket. The observer, 

 looking through the eyepiece which receives 

 a portion of the light from a star, makes 

 the exposures while a recorder in another 

 room reads the galvanometer. As this sec- 

 ond room may be heated, it is our custom, 

 especially in winter, to reverse astronomical 

 practise by having the chief observer write 

 down the notes, while the assistant is sent 

 up into the cold dome to manipulate the 

 telescope. 



There is another device, however, which 

 bids fair to supplant entirely the selenium 

 photometer, namely, the photo-electric ceU 

 made from one of the alkali metals. The 

 sensitive metallic surface is in an exhausted 

 tube with a small quantity of inert gas, 

 and the effect of light is to release electrons 

 from the surface, which ionize the gas, and 

 thus a current is produced. We are fortu- 

 nate in having several of our physicists at 

 Illinois interested in photo-electric cells, 

 especially Professor Jacob Kunz, and it is 

 in the laboratory where the really impor- 

 tant improvements are made. Only re- 

 cently we managed to produce a cell which 

 is twice as sensitive as anything we had 

 before, and this amounts to the same thing 

 as though some good fairy had suddenly 

 doubled the light gathering power of our 

 telescope. The great advantages of the 

 photo-electric cell over selenium are first 

 the freedom from irregularities, and next 

 the very short time of recovery. It is too 



