MOUNT WILSON OBSERVATORY. 249 



discharge outside the block of wood; and as only light from the slot passes 

 through the slit of the spectroscope, the resulting spectrogram shows httle or 

 nothing. If the wires are slightly amalgamated or alloyed with some other 

 metal, they explode in the normal way, and then the spectra of copper, silver, 

 or gold are of about the same intensity as those of other metals. This sug- 

 gests that the vapors of copper, silver, or gold in the pure state are very poor 

 conductors of electricity. All of the other metals tried behave normally. 

 The absorption lines are in general those which show regularly in the arc 

 spectrum of the element. Because of great intensity in the ultra-violet and 

 the relatively small number of absorption lines, the explosions of aluminum, 

 zinc, and lead are especially well adapted to serve as sources for the study of 

 absorption in this part of the spectrum. 



Spectral Energy Distribution. 



With the aid of a vacuum thermopile, made by Messrs. Nicholson and 

 Pettit, and a galvanometer, in conjunction with a quartz spectrograph, some 

 of the preliminary work was done on the mapping of the energy spectrum of 

 the explosions of iron wires. Easily measurable deflections were obtained 

 from about X = 2 n in the infra-red to X = 1990 a in the ultra-violet. The 

 object of these experiments was to see if energy measures are possible with 

 this type of source, and to secure data which will be needed for the design of 

 apparatus suitable for accurate work of this kind. They show that a quanti- 

 tative study of the radiation of metallic vapors at very high temperatures is 

 not only possible, but does not seem to offer any very great difl&culties. 



Opacity of Vapors at High Temperatures. 



This quantity, which plays an important part in theoretical investigations 

 in astrophysics, may be investigated directly with the explosions. By passing 

 the hght from one explosion (A) through a second one (B) connected in series 

 with it, it is possible to measure how much of the light of any particular wave- 

 length from A passes through B. It is found that if B is an explosion in a slot, 

 giving a continuous spectrum, while A is a spark or an explosion in the open 

 giving a discontinuous spectrum, B is quite opaque, no light from A being 

 transmitted through it; while, if the light from B passes through A, a con- 

 siderable fraction is transmitted, showing that A is only partially opaque. 

 The method, when properly developed, promises to give quantitative data 

 concerning the absorption coefficient, especially when used in conjunction 

 with the method described in the next paragraph. 



Velocity of Sound in High-Temperature Vapors. 



A method has been developed which makes it possible to determine the 

 velocity of sound in metallic vapors at very high temperatures. This method 

 is as follows: An image of the open-air explosion seen end on, for example, is 

 projected upon the slit of a rotating-mirror camera. The photograph made in 

 this way shows the explosion drawn out into a band of light of great 

 intensity at the beginning, but gradually fading out as the vapors cool by 

 radiation. The image of the slit moves over the photographic plate at the 

 rate of about 40,000 cm. per second, and as the records obtained for different 

 elements vary from 8 to 20 cm. in length for open-air explosions, it follows 

 that the vapors are luminous during — ^ to ^^ of a second. The first 3 or 



