542 Sir E. Rutherford on Collision of 



the H scintillations appeared larger and more diffuse, probably 



due to the scattering of" the light in passing through the thick 



la3 r er of crystals, and were more easily counted, while weak 



scintillations could be counted on a brighter background than 



with the ordinary screen. At the same time, the layer of 



crystals was so uniform, that each incident H atom produced 



a scintillation. 



In these experiments, two workers are required, one to 



remove the source of radiation and to make experimental 



adjustments, and the other to do the counting. Before 



beginning to count, the observer rests his eyes for half an 



hour in a dark room and should not expose his eyes to any 



but a weak light during the whole time of counting;. The 



experiments were made in a large darkened room with 



a small dark chamber attached to which the observer retired 



when it was necessary to turn on the light for experimental 



adjustments. It was found convenient in practice to count 



for 1 minute and then rest for an equal interval, the times 



and data being recorded by the assistant. As a rule, the eye 



becomes fatigued after an hour's counting and the results 



. . . 



become erratic and unreliable. It is not desirable to count 



for more than 1 hour per day, and preferably only a few 



times per week. 



Under good conditions, counting experiments are quite 



reliable from day to day. Those obtained by my assistant 



Mr. TV. Kay and myself were always in excellent accord 



under the most varied conditions. It was usually arranged 



that the number of scintillations to be counted varied between 



15 and 40 per minute. 



§ 5. Experimental arrangement. 



For experiments with hydrogen and other gases, the active 

 disk D (fig. 1) was mounted at a convenient height parallel to 

 the screen on a metal bar B which slid into a rectangular brass 

 box A, 18 cm. long, 6 cm. deep, and 2 cm. wide, with metal 

 flanges at both ends fitting between the rectangular poles of 

 a large electromagnet. One end was closed by a ground glass 

 plate C, and the other by a waxed brass plate E, in the centre 

 of which was cut a rectangular opening 1 cm. long and 3 mm. 

 wide. This opening was covered by a thin plate of metals 

 of silver, aluminium or iron, whose stopping power for 

 a particles lay between 4 and 6 cm. of air. The zinc sulphide 

 screen F was fixed opposite the opening and distant 1 or 2 mm. 

 from the metal covering. By means of two stopcocks, the 



