October 20, 1922] 



SCIENCE 



453 



short atmospheric temperature changes from 

 without, such as would not be otherwise re- 

 corded, may make an impression on the 8 P.M. 

 graph. This, however, would not bear upon 

 the 1922 graph as a whole, from July 24 to 

 August 14. Supposing, moreover, that the 

 closed region within is in some way modified 

 thermally by the high exhaustions (carried to 

 within .001 mm.), it seems hardly probable 

 that the apparatus would take so long to return 

 to the normal condition of 1921. 



What has gone down during this series of 

 measuremeents is the vacuum and one would 

 therefore conclude that "states of high exhaus- 

 tion (a feAv hundredths or tenths of a mm.) 

 are (like the plenum) more susceptible to the 

 presence of radiant activity than the lower ex- 

 haustions of a few mm. Thus, night observa- 

 tions presupposed, the radiant forces pass 

 through a minimum in a partial vacuum, of 

 several millimeters or more, and the best con- 

 ditions for observation are then at hand. To 

 test this further, I exhausted the apparatus on 

 August 14. Tlie morning observations August 

 15, twelve hours later (see figure) are again 

 abnormally high. 



It not infrequently happens that night 

 values are low when day values are high and, 

 in general, there is a tendency of the graphs 

 to converge toward rainy or densely cloudy 

 weather. All this conforms with the view that 

 the needle is screened from radiation by the 

 large attracting mass BI and that the radiant 

 forces act with gravitation, if the temperature- 

 time coefficient db/dt is positive, and act 

 against gravitation when d^/dt is negative, as 

 elsewhere explained. I have been tempted to 

 envisage a coeflicient dQ/dt, which is not all 

 temperature ;, for there may be some other radi- 

 ation or agency behind the recent rains (for 

 instance), as well as behind the diiierence in 

 the character of the results of 1922 and 1921 

 as exhibited by the figure. It is difficult, in 

 other words, to surmise what the nature of the 

 radiant discrepancy may be, which clings to 

 the apparatus so persistently in July and early 

 August. If it were merely thermal, or de- 

 pendent on a kinetic mechanism associated 

 with d^/dt, its behavior would seem to be 

 incompatible with the daily cycle, which is 



practically immediate. Hovrever, if the slopes 

 of the curves giving the static elongations, y, 

 of the needle in the lapse of time, are en- 

 hanced by the higher degrees of exhaustion, 

 these curves would also ultimately intersect, 

 so that even negative values of Aj/, referable 

 to causes within the apparatus would not be 

 unexpected. 



On my return to the laboratory in Septem- 

 ber, I resumed the work (upper curve). The 

 vacuum had in the mean time decreased to 

 about 3 mm. Under these conditions the night 

 observations (8^) are again normal and com- 

 IJare favorably with the corresponding graph 

 of 1921, as was anticipated. 



Carl Baeus 



Bbown University, 

 Providence, E. I. 



THE AMERICAN CHEMICAL 

 SOCIETY 



(Continued) 



Division op Dte Chemistry 



William J. Hale, chairman 



E. Norris Shreve, secretary 



symposium on methods for standardizing and 



testing dyes 



B. E. Eose, chairman 



■ Introductory remarlcs: Eobert E. Eose. 



Chemical control of dyestuffs: Walter M. 

 Scott. This paper presents a general diseussiou 

 of various methods for estimating the strength 

 of dj'estuifs as follows : ( 1 ) Colorimetric com- 

 parison of standard dyestuff solutions. (2) Titra- 

 tion of a solution of known strength of dyestufE 

 with a standard solution of titauous chloride in 

 an atmosphere of carbon dioxide. (3 Determina- 

 tion of the percentage of nitrogen by the Kjeldahl 

 method. (4) Estimation of the inorganic salts 

 which have been used in the standardization of 

 the dyestuff. In connection with the materials 

 used in dyeing there is such a great variety that 

 it is only possible to discuss a few of the more 

 common types. This paper gives an outline of 

 the general methods of analysis used and also 

 suggested specifications for the following: acetic 

 acid, sulfuric acid, ammonia, black iron liquor, 

 commercial ' ' nitrate ' ' of iron, di-sodium phos- 

 pliate, Glauber's salt, common salt and soap made 

 from olive or red oil. 



The estimation of erythrosine : W. C. Holmes. 

 A method is outlined for the direct evaluation of 



