46 Scientific Proceedings, Royal Dublin Society. 



the real staudard adopted in comparing the various figures. ^ In the case of very 

 thin screens the unavoidable variation in thickness of wall of different emanation 

 tubes would cause appreciable errors, but for brass screens thicker than 0-2 mm. 

 this variation would be unimportant. 



In each of the four tables the action is shown at various points on a plane 

 parallel to the tube or tubes. In Tables 1 and 3 this plane passes through the 

 axis of the tube. In Tables 2 and 4 it bisects the rectangle formed by the tubes 

 at right angles. The relations of the vai'ious planes and distances involved may, 

 perhaps, be described as follows : — 



Suppose that the emanation tube or tubes are placed horizontally in a north 

 and south direction. Tables 1 and 3 refer to short and long single tubes respec- 

 tively, while Tables 2 and 4 refer respectively to four parallel short lubes and six 

 parallel long tubes, covering in each case a horizontal rectangular area. What 

 has been called the " equatorial plane^" is now a vertical east and west plane, 

 pa.ssing through the centre of the tube or tubes. The figures in the tables refer 

 to points in a vertical north and south plane, which, in the case of Tables 1 and 3, 

 passes through the axis of the tube, and, in the case of Tables 2 and 4, passes 

 midway between the two central tubes of the area: a in the tables is the vertical 

 height of the given point above or below the tube or tubes ; 5 is the horizontal 

 distance of the given point north or south of the centre of the tube or tubes. 



In the case of a single tube [Tables 1 and 3] the distribution is obviously 

 cylindrical round the axis of the tube, so that the action at any point may be 

 found from the tables. It is only necessary to choose as the plane of the table 

 one which passes through the axis of the tube and the required point, and measure 

 a and h in this plane. In Table 2 the effective area covered by the four tubes is 

 approximately square, so that the distribution in the "east and west" or '"equa- 

 torial " plane is almost identical with that in the " north and south " plane, at 

 distances greater than 5 nrm. from the tubes. In Table 4 the effective area 

 covered by the tubes may be taken as 100 mm. x 60 mm., so that 30 nmi. east or 

 west of the centre brings us to the edge, whereas 50 mm. is required to do so in 

 the plane of the table. It would appear then that, if we wish to find the distri- 

 bution in the equatorial plane, no large error would be introduced by taknig from 

 the table a value for the action at a point corresponding to a value of I 20 mm. 

 greater than the actual distance east or west of the centre. This would only 

 apply to points at distances of 10 mm. or more from the plane of the tubes. 



T'be figures in the second column of Table 1 correspond to those in column A 

 of the table in the previous paper. The slight discrepancies observable in some 

 cases are due to the use of a slightly different absorption coefficient for -y rays in 

 flesh, as a small discrepancy had occurred between the values used for column A 

 and those employed in columns B, C, and D of the previous table. Results are 

 shown to two significant figures only. 



Some examples on the reading of the tables may be of use. Suppose a small 

 tumour is to be treated, and it is decided to use an arrangement of needles similar 

 to that given for Table 2. If the total dose distributed over tlie four tubes is 

 50 millicurie hours, tlie action to be expected would be : — 



On the surface of a serum needle . . 50 x 1600 = 80,000 units. 



At a depth of 1 cm. opposite to centre of tubes 50 x 9"1 = 455 ,, 



ends „ 50 x 6-9 = 345 „ 



5 cms. „ centre „ 50 x 0-37 = 18-5 „ 



ends „ 50 x 0-37 = 18-5 „ ■ 



