Measurement of Electromagnetic Radiation. 59 



While this very small degree of heating was within the 

 rano-e of experimental certainty so long as the clockwork was 

 goin°", it was impossible to keep any index visible at all when 

 the clock had stopped, even though several concentric screens 

 were placed round the tube. 



The apparatus in this form was subject to the great fault 

 that the phosphorus refused to form the oxide at a temperature 

 below 15° C, which prevented the instrument being used on 

 many days. 



With the view of getting over this difficulty with the 

 phosphorus index, and if possible of obtaining greater sen- 

 sitiveness, a form of index was adopted which was a modifi- 

 cation of Dr. Joule's arrangement. 



The tube was closed at each end, the partition reaching 

 right up to the cork at the top and leaving a space of about 

 fifteen centimetres at the bottom. Twelve centimetres below 

 the top of the partition a hole was punched in the card. Into 

 this hole a short cylinder of paper was glued. This is shown 

 at A in figs. 8 & 9. A plane mirror, M, fitted into this tube, 

 allowing about 1 millimetre clearance all round. It was 

 fastened to a cross bar, at the other end of which a counter- 

 poise (W) was fixed. 



The mirror and counterpoise were suspended by a fine quartz 

 fibre from a pin (E, fig. 8), which passed through the cork 

 used to close the tube. When the mirror is at rest it almost 

 completely fills up the hole in the partition, and forms a very 

 delicate means of detecting any differences of pressure in the 

 air at the two sides of the partition. 



A small recess (B, fig. 9) was made in the partition for the 

 counterpoise, and covered in so that it was not exposed to the 

 current of air. The arrangement is shown in elevation 

 in fig. 8, and in section through KL in fig. 9. Fig. 10 

 shows the mirror and counterpoise separately. A hole was 

 ground in the tube immediately in front of the mirror, into 

 which a lens of a metre focal length was cemented. 



In this form of the instrument the guard-tube, if continuous, 

 would have prevented the formation of a clear image. It 

 was therefore necessary to divide it into two portions. Rings 

 of mahogany (N, fig. 11) were cemented to the top of 

 the lower portion and to the top and bottom of the upper. 

 The two rings N x N 2 were connected together by three bolts, 

 of such a length as to leave a space of about two centimetres 

 between the rings. This break in the guard-tube came just 

 on a level with the mirror, and thus allowed the beam of light 

 to pass freely. The passage of the bolts before the mirror 

 produced no inconvenient effect on the image formed on the 



