May 21, 1897.] 



SCIENCE. 



795 



The plate to be discharged was then 

 placed in the shadow of an opaque obstacle, 

 and raj's allowed to strike only the air near 

 the plate. The form of the curve was not 

 different from that obtained when the rays 

 were partially screened off from both the 

 plate and the surrounding air. These two 

 experiments do not corroborate the theory 

 of Perrin as given above. 



It is also to be noted that the conditions 

 in this last experiment were similar to those 

 under which Perrin worked when he found 

 the relations between the rate of discharge 

 and the density of the gas, but the results 

 which I found did not agree with the law 

 as given by him. 



There is also a fact in connection with 

 this work which is worthy of note, although 

 it does not directly bear upon the experi- 

 ments here described. After the discharge 

 had taken place the plate would often ap- 

 pear to recharge to a very noticeable degree. 

 The electrometer used was a ' dead beat ' 

 electrometer. The recharging seemed to 

 be more noticeable when the original poten- 

 tial was small, and the intensity of the rays 

 great. My apparatus is not well adapted 

 to study this phenomenon, nor have I had 

 the time to do so. I hope hereafter to give 

 a better proof of the existence of this phe- 

 nomenon than I can at present offer, and to 

 study the conditions under which it occurs 

 more fully. 



The experiments which I have described 

 indicate a conduction eifect through the gas 

 rather "than a convection effect due to parti- 

 cles thrown off from the discharging plate, 

 but it would seem scarcely advisable to at- 

 tempt to form a theory to explain these 

 phenomena until they have been studied 

 more completely. 



Since writing the preceding I have inves- 

 tigated more completely the dependence of 

 the rate of discharge caused by X-rays upon 

 the potential used at different pressures of 



the surrounding gas. I find that the limit- 

 ing value of the current is sooner reached 

 when the experiment is carried on at low 

 pressures than it is at higher ones, and that 

 the limiting values, called by Thomson and 

 Rutherford saturation points, are roughly 

 proportional to the square roots of the den- 

 sity of the gas. 



Stoletoiv iJourn. de Phys. 9, 471 ) found 

 that in the case of discharge caused by 

 ultra-violet light the pressure of the gas at 

 which a maximum effect occurred was 

 proportional to the potential with which 

 he was working. I have tested the dis- 

 charge caused by the X-rays in this 

 respect, and for this purpose I used much 

 greater intensity of radiation than I had 

 previously used. I found the pressure for 

 maximum effect to be roughly proportional 

 to potential of the charged plate. Also 

 the intensity of the rays has a very great 

 effect on the point of maximum effect. The- 

 greater the intensity the lower was this 

 point. 



I have also tried allowing the rays to 

 strike the charged body at normal incidence; 

 and the results were the same which I had 

 previously found when the incidence was 

 grazing. 



C. D. Child. 



COENELL UNFVEESITY. 



CUBBENT NOTES ON PHYSIOGBAPHY. 

 CORNISH ON SAND DUNES. 



Mb. Vatjghan Cornish discusses the 

 formation of sand-dunes (^London Geogr. 

 Journ., IX., 1897, 278-309), and throws 

 much light on their growth and movements. 

 Basing his work on observation and experi- 

 ment he discusses the effects of supply 

 and texture of sand and of direction 

 and strength of wind, and reaches satisfac- 

 tory explanations of transverse, longitudinal 

 and crescentic dunes {harchanes of Arabia, 

 medanos of Peru), adding a suggestive 

 hypothesis for the origin by wind-excava- 



