306 Prof. R. TV. Wood on Uneven Distribution of 



balance. However the well-defined chemical properties of 

 both ThX and UrX are not in accordance with the view that 

 the actual amounts involved are of this extreme order of 

 minuteness. On the other hand, the existence of radioactive 

 elements at all in the earth's crust is an a priori argument 

 against the magnitude of the change being anything but 

 small. 



Radioactivity as a new property of matter capable of exact 

 quantitative determination thus possesses an interest apart 

 from the peculiar properties and powers which the radiations 

 themselves exhibit. Mme. Curie, who isolated from pitch- 

 blende a new substance, radium, which possessed distinct 

 chemical properties and spectroscopic lines, used the property 

 as a means of chemical analysis. An exact parallel is to be 

 found in Bunsen's discovery and separation of cpesium and 

 rubidium by means of the spectroscope. 



The present results show that radioactivity can also be 

 used to follow chemical changes occurring in matter. The 

 properties of matter that fulfil the necessary conditions for 

 the study of chemical change without disturbance to the 

 reacting system are few in number. It seems not unreasonable 

 to hope, in the light of the foregoing results, that radioactivity, 

 being such a property, affords the means of obtaining infor- 

 mation of the processes occurring within the chemical atom, 

 in the same way as the rotation of the plane of polarization 

 and other physical properties have been used in chemistry 

 for the investigation of the course of molecular change. 



Macdonald Physics Building, 



Macdonald Chemistry and Mining Building, 



McGill University, Montreal. 



XLII. On a Remarkable Case of Uneven Distribution of Light 

 in a Diffraction Grating Spectrum. By R. TV. AVood, Pro- 

 fessor of Experimental Physics, Johns Hop)hins University*. 



IT is a well-known fact that in the spectra formed by a 

 diffraction-grating the light is unevenly distributed, that 

 is the total light in any one spectrum will not recombinc to 

 form white light. 



I have been examining a most remarkable grating recently 

 ruled on one of the Rowland dividing-engines in which this 

 uneven distribution is carried to a degree almost incompre- 

 hensible. If the spectra of an incandescent lamp are viewed 

 directly in the grating without any other optical appliance,, 

 at certain angles of incidence perfectly sharp monochromatic 

 * Communicated by the Physical Society: read June 20, 1902. 



