98 ROYAL SOCIETY OF CANADA 



There again it is seen that the elt'eet of the secondary radiation 

 was quite marked. As the results show the ionization steadily de- 

 creased to a minimum value which corresponded approximately to the 

 position in which the radium was on the line joining the electroscope 

 to the corner of the room. After passing through this minimum value 

 the ionization then steadily increased and reached a maximum when 

 the radium was slightly beyond a line drawn from the cylinder perpen- 

 dicular to the far wall. After this the ionization fell away again as 

 the distance of the radium from both walls was increased. 



The maximum variation in the values of " n "' for these experi- 

 ments, it will be seen from Tables YII and VIII amounted to as much 

 as 10 per cent. It seemed reasonable, therefore, to suppose that the 

 presence of some such object as a brick wall might, when the penetrating 

 rays from the earth impinged upon it, in the same way affect the natural 

 ionization in any metallic cylinder, and care was, therefore, taken in 

 the measurements made when determining the screening effect of the 

 lake and of different soils, to place the electroscope and receiver as far 

 away as possible from any building which might modify in some such 

 way as that indicated the ionization of the enclosed air. 



V. — Screening Experiments. 



Being assured from the foregoing experiments that changes in the 

 ionization due to daily variations were inconsiderable in comparison 

 with variations due to a change in position, three series of measure- 

 ments were then undertaken with the object of investigating the screen- 

 ing effect of Lake Ontario, care being taken to choose positions of 

 observation as far as possible from any artificial surroundings. The 

 first and second series were made with lead cylinders after the ionization 

 liad reached the steady state and the third with well cleaned cylinders 

 of lead, zinc and aluminium containing freshly filtered air. A con- 

 siderable decrease in the ionization when measured over the water was 

 noticeable in every case, a reduction of as high as 60 per cent being 

 recorded in the case of the freshly cleaned lead cylinder in the third 

 series of measurements referred to above. 



1. — Measurements on hoard steamer " Corona." 



In the first set of measurements a series of observations was made 

 on a selected day in the laboratory on the conductivity of the air en- 

 closed in a lead receiver which had not boon recently cleaned, and the 

 mean of these readings was found to give a value of 42 ions per c.c. 

 per sec. for the conductivity. Measurements wore then made on the 

 same day on board the SS. " Corona " during one of her voyages, and 



