APPENDIX. 



NATURAL ION1SATION IN CLOSED MKTAL CONTAINERS. 



During the voyage of the " Terra Nova " from Cardiff to Lyttelton, New Zealand, 

 observations of the number of ions formed per second within a vessel of about 

 27,000 c.cs. capacity were regularly made and the results have been published in the 

 ' Proceedings of the Royal Society/ A, vol. LXXXV, 1911. 



The most important results were to show that the number of ions formed 

 per c.c. per second in such a vessel was lower than the number formed on land : that 

 a considerable addition to the normal ionisation over the sea might, under favourable, 

 circumstances, be due to radioactive products deposited on the ship and due primarily 

 to a high radium content of air derived from neighbouring land masses ; moreover, as 

 in previous experiments on Lake Ontario,* that the value over the sea could not be 

 reduced below a certain limit in this case 4 ions formed per c.c. per second. A mean 

 value of 3-8 ions has also been observed on the fourth cruise of the " Carnegie." 



The question of the lower limit which can be reached is of importance, since the 

 cause of the residxial ionisation is still obscure. For this reason a few measurements 

 usino- the apparatus and method described in the paper cited above were made in 1911 

 within a cave excavated in a snowdrift consolidated to ice. In this cave the vessel was 

 screened by many feet of ice on all sides, except from radiation coming from above 

 and from the earth. Three feet of ice separated the cylinder from the upper surface 

 of the drift, and one foot of ice from the ground below. It is unfortunate that the 

 opportunity was not taken to measure the radium content of the ice in order to determine 

 whether there was any notable contribution to the ionisation from this cause. t The 

 radium content of the rock was not ascertained, but the surface of the ground had for 

 many years been protected from the deposition of radioactive substances by the ice 

 covering above it. It was thought that, in these surroundings, the residual ionisation 

 ini^lit be even less than that which was observed 011 the " Terra Nova," and this proved 

 to be the case. 



The lowest value observed on 14 days in July, 1911, was in fact just over 3 pairs 

 of ions formed per c.c. per second, though considerable variations were observed from 

 day to day. The mean value during the period was almost exactly the same as the 

 lowest observed on the " Terra Nova," viz., 4 pairs of ions formed per c.c. per second. 



* Wright, ' 1'liil. May.,' February, 1 !)()'.) (lowest value (>(> inns). A later determination by .McLennan, 

 usinu a cylinder of ice over Lake Ontario, ave a value of only 2-0 ions. 



| An indication of climatic variations mi-rlit well he obtained by systematic investigation of the 

 radioactive content of glaciers at different depths. 



43 



