496 



Prof. C. V. Boys. 



[Apr. 24,. 



that the radio-micrometer will respond to a temperature rise of less' 

 than one-millionth of a degree. Even if my figure of one-millionth is- 

 not absolutely correct — and I believe that the figure should be still 

 smaller — the discrepancy is so enormous as to require explanation- 

 Knowing how completely every thermopile that I have examined fails 

 to approach in delicacy the radio-micrometer, I cannot quite under- 

 stand how Mr. Stone was able to observe deflections which correspond 

 to a rate of radiation which is only a few times as great as the least 

 that I can observe. As an instance of the delicacy of the radio- 

 micrometer, I may state that, compared with the usual form of ther- 

 mopile, made by Messrs. Elliot Brothers, connected with a low 

 resistance reflecting galvanometer by the same makers, it is fully 

 one thousand times as sensitive, besides being far more stable or 

 insensitive to disturbing causes. 



I think my observations go to show that the heat of Arcturus has 

 not as yet been observed (unless the refracting telescope possesses 

 some mysterious power which the reflector does not) ; and the same- 

 conclusion may be drawn with respect to the other stars. I have by 

 no means reached what I believe to be a practical limit to the 

 delicacy of the radio-micrometer, and it is possible that with a more 

 delicate instrument, or with a larger telescope — and Mr. Common- 

 has promised to allow me to use for the purpose his 5-foot reflector 

 — I may jet be able to observe some definite and real effect ; but that 

 I should have so far failed is not surprising, in view of the following 

 argument. "When anything is heated and gives out light an incre- 

 ment of temperature causes a relatively greater increment of light ; so' 

 that, light for light, less heat is radiated by the hotter body. Thus,, 

 among stars that are equally bright, the coolest will send the most 

 heat. Now, a candle-flame is, in all probability, far cooler than 

 any white or yellow star ; and so if a candle-flame is removed until 

 it appears of the same brightness as some star — for instance, Arc- 

 turus — then it should radiate far more heat. Now, I do not know 

 at what distance a candle would appear as bright as Arcturus,, 

 but supposing that it is as much as 171 miles, then, since at this, 

 distance I can only just detect the heat of the candle, it is evident 

 that I should not be able to detect the heat of Arcturus. 



With regard to the observations on the moon, they are, I know,, 

 exceedingly fragmentary and incomplete. They were not, however, 

 undertaken for their own sake, though they may have some value, but 

 in order to test the apparatus and to serve as a standard by which 

 other observations might be compared with mine. For work on the 

 moon the apparatus is not exactly the most convenient, though it does 

 well. For systematic work I should employ the most delicate circuit* 

 I could make with a very small sensitive surface and a quick period, 

 and then, to obtain curves of heat which should give the deflections- 



