784 



SCIENCE 



[N. S. Vol,. XXVII. No. 



■west winds. Between two and three o'clock 

 in the afternoon, about three hours after the 

 fire started, the updraft was sufficiently strong 

 to overcome the high wind and occasionally 

 to carry water vapor to the level of cloud 

 formation. Cumulus clouds resulted, capping 

 the smoke, and appearing or disappearing ac- 

 cording as the latter rose or failed to rise to 

 the necessary altitude. So near to this alti- 

 tude was the average summit of the smoke 

 that it was possible for the writer, on seeing 

 an especially vigorous puff from the fire, to 

 predict the formation of a cloud some seconds 

 in advance of its appearance. The clouds did 

 not, as far as could be seen from a position 

 directly to windward of the fire, attain to well- 

 rounded, typical cumulus forms. They varied 

 from mere flecks of white to moderately large 

 but flattish masses and were usually dissipated 

 within five minutes from the time they became 

 visible. Their bases were more or less mingled 

 with and hardly distinguishable from the sum- 

 mit of the smoke-cloud; it was therefore im- 

 possible to tell whether or not they were typic- 

 ally flat-based. The clouds appeared to be 

 formed not directly over the fire, but a very 

 considerable distance to leeward, where the 

 high wind first permitted the rising air to 

 reach its dew-point altitude. 



The occurrence of these cimiuli recalls a 

 similar phenomenon over the burning coal 

 pockets of the Boston & Maine Railroad close 

 by at Charlestown in December, 1896, and 

 noted by Professor E. DeC. Ward in SciENCfE 

 for January 8, 1897. In this instance the 

 greater concentration of the fire and the con- 

 sequent greater proportion of water-vapor car- 

 ried aloft, caused the development of a far 

 more perfect cumulus cloud than that formed 

 over the widely scattered Chelsea fire. 



B. M. Varney 



Habvabd Univeksitt, 

 April 30, 1908 



THE INFALLIBILITY OF NEWTON's LAW OF 

 RADIATION AT KNOWN TEMPERATURES 



Although there is no direct reference to 

 " the absolute temperature of space " (on 

 which hinges the whole question of the sun's 

 effective surface temperature) in Professor 



Verys paper published in the last number of 

 Science, it is clear that he is still inclined to 

 favor the claim that the temperature of space 

 is in the neighborhood of 300° C, notwith- 

 standing the demonstration I have given, 

 showing that the temperature is probably less 

 that 2° C. 



The title of the present article gives evi- 

 dence that I wholly disagree with Professor 

 Very when he claims that Stefan's law is in 

 better agreement with actual observation than 

 is Newton's law. 



Just why I regard Stefan's law as wholly 

 wrong will appear from, the theoretical results 

 given below. How such erroneous laws simi- 

 lar to that of Stefan's ever came to be de- 

 duced can be largely inferred from the con- 

 tents of a paper on " The Earth as a Heat- 

 radiating Planet," sent to the editor of Sci- 

 ence on December 25, 1907, but not yet pub- 

 lished at the time of this writing.' In that 

 paper (where, for obtaining the terrestrial 

 radiation into space, the effective surface- 

 temperature of the earth is provisionally 

 placed at 200° C.) it is made evident that ac- 

 cording to my results " serious changes in the 

 constants of radiation in the formulae accepted 

 to-day " (to quote part of a sentence from 

 Professor Very's article) must actually be 

 made. 



I shall now demonstrate that both theo- 

 retically and experimentally Newton's law 

 gives uniformly consistent results when the 

 observations are properly interpreted, and that 

 Stefan's law leads to absurd and unintelligible 

 results at known temperatures. 



Let us first conceive that the observations 

 were made in free space, the two totally dif- 

 ferent expressions for the absolute tempera- 

 ture of space will then read 



For Newton's law t = T{d/D)' = 0°.7 

 For Stefan's law t = T yj d/D = 300°. 



Since the temperature of space must be taken 

 as constant in each case we obtain for com- 

 parison the two sets of values of T, for dif- 

 ferent values of D, given in the second and 

 third columns of the following table: 



■ Published in Science for March 6, 1908. 



I 



