Radiation from the EartKs Surface. 409 



when it was left exposed in the laboratory, it follows that 

 the bricks must have supplied an amount of penetrating 

 radiation equal to the amount they cut off. This result 

 mioht have been anticipated, as the walls of the laboratory 

 were of brick. 



The activity of lead slabs which had been exposed for over 

 two weeks on the roof of the buildino- was tested, and also 

 the activity of the copper roof itself. Both exhibited greater 

 activity than materials which had been kept in the laboratory, 

 but the radiation was not of a penetrating character. Ex- 

 periment showed that the amount of penetrating radiation 

 was approximately the same in the laboratories, on the roof, 

 and on the bare ground outside the building ; in fact this 

 radiation seems to be present everywhere with practically 

 uniform intensity. 



The effect of placing metal sheaths as linings inside the 

 electroscope was observed in an attempt to reduce the ioniza- 

 tion as much as possible, and also the instrument with its 

 base removed was placed on other non-metallic substances. 

 In this way dried earth, polished tin, writing-paper, a stone 

 window-ledge, pine wood, brick, lead, zinc, and iron were 

 tried ; the results showing that they were all, with the ex- 

 ception of iron, more active than polished brass. Iron exhibited 

 about the same effect as brass, but it is probably more active, 

 for the iron used was a slab about '75 of an inch in thickness, 

 which must necessarily have absorbed a considerable amount 

 of the penetrating rays. Great care was not taken in this 

 part of the work to secure numerical values for the activity 

 of the various substances, as the object was merely to deter- 

 mine if any material showed a lower ionizing effect than 

 brass. The results would, however, indicate that the above 

 list represents the materials arranged according to their 

 activity, beginning with the most active. The quantitative 

 aspect of this branch of the subject has been carefully in- 

 vestigated by McLennan and Burton^, and by R. J. Struttf. 



In order to calculate the number of ions produced per c.c. 

 per sec. it was necessary to reduce the readings on the micro- 

 meter scale to volts, to estimate the electrical capacity of the 

 insulated gold-leaf system, and to calculate the volume of the 

 inclosed air. 



The deflexions of the gold leaf in micrometer divisions 

 could be reduced to readings in volts by observing the de- 

 flexions when the leaves were charged to different known 



* McLennan and Burton, Phys. Rev. Ixxxiv. p. 184. 

 t R. J. Strutt, ' Nature,' Feb. 19, ] 903, p. 369 ; see also ' Nature,' 

 1903, pp. 391, 414, 439. 



