480 



SCIENCE. 



[N, 3. Vol. VII. No. 171, 



Oa the same scale a represents the cooling 

 from 9 degrees to 6 degrees ; aa from 6 de- 

 grees to 4 degrees, and aaa from 3 degrees 

 to 2 degrees. Now, Newton's law of cool- 

 ing requires that the rate shall vary directly 

 with the difference of temperature between 

 the cooling body and the surrounding me- 

 dium. While this law is known to be in- 

 correct for large differences of temperature, 

 it is generally accepted for very small differ- 

 ences. If it were correct under the condi- 

 tions of the present experiment, then the 

 ratios of the times required for the temper- 

 ature to fall through the several ranges 

 above indicated would all equal unity, and 

 the curves A, a, aa, aaa would coalesce. 

 But they are very far from doing this. It 

 will be observed that all of these curves 

 preserve their relative values very closely 

 indeed, until they approach the point of 

 pressure where the curve A reverses itself; 

 then they begin to bunch themselves very 

 much closer together, especially the lower 

 ones, and shortly reach a greatly reduced 

 as well as varied ratio of values which they 

 retain substantially unchanged to the end, 

 as shown in connection with section C. To 

 avoid confusion of lines, I have omitted the 

 secondary curves corresponding with sec- 

 tion B. 



Carbon monoxide was chosen for com- 

 parison with air, because its absorptive 

 power for radiant heat is many times 

 greater, while its specific heat is almost ex- 

 actly the same. The principal curve, repre- 

 senting the rate of heat transmission from 

 15 degrees to 10 degrees, differs very little 

 from that of air. It shows a slightly better 

 rate than air at very small pressures ; not 

 quite so good a rate as air at intermediate 

 pressures ; and the same rate at atmos- 

 pheric pressure. But the curves a, aa, 

 aaa, i-epresenting equivalent amounts of 

 cooling at smaller temperature differences, 

 are materially unlike those of air. At high 

 pressures they have about the same ratio 



values as with air ; but the ratio diminishes 

 much less at intermediate and low pressures; 

 that is to say, the curves remain further 

 apart. It is equally noticeable that the 

 curves aa, aaa retain their full relative 

 ratio values at low pressures, while with 

 air they nearly coalesce. 



It was thought that ethylene might trans- 

 mit heat more rapidly than air, because of 

 its much higher specific heat. But it does 

 not do so. Its curve has the same form 

 as those of air and carbon monoxide. It 

 transmits heat nearly as well as air at at- 

 mospheric pressure, but not nearly so well at 

 intermediate pressures. At a very few mill- 

 ionths, however, it conducts a trifle better 

 than air. The curves a, aa and aaa have 

 the same characteristics and about the 

 same ratios as those of carbon monoxide. 



Hydrogen was next tried, on account of 

 its very low coef&cient of viscosity, as well 

 as its very high specific heat. While in 

 general form the hydrogen curve resembles 

 the air curve, all the ordinates are im- 

 mensely increased. It is noticeable that 

 the intermediate section B of the curve lies 

 much nearer A than C, quite different from 

 its relative position in the curves of the 

 other gases. This section of the curve 

 shows that hydrogen retains about two- 

 thirds of its initial heat transmitting power 

 at a pressure nearly two hundred times 

 smaller than does air. The curves A, a, 

 aa and aaa have something like the same 

 ratios as they have in the cases of carbon 

 monoxide and ethylene. In general, it may 

 be said of hydrogen in the large radiation 

 bulb, that it transmits heat nearly four 

 times as fast as air at atmospheric pressure ; 

 more than twice as fast at a very few mill- 

 ionths, and more than seven times as fast 

 through a long range of intermediate pres- 

 sures. 



As evidence of the accuracy of the ob- 

 servations on which the curves thus far 

 described are based, it is gratifying to note 



