286 



SCIENCE. 



[N. S. Vol. XVI. No. 399. 



perature in one and a simultaneous lower- 

 ing in the other vessel. Such an explana- 

 tion, however, is unsatisfactory. How can 

 elastic pai'ticles (molecules) do work on 

 elastic particles? From the impacts of 

 elastic molecules at a given temperature but 

 with all degrees of velocity, there can result 

 nothing other than all degrees of velocity 

 with the same temperature (only the mean 

 free path would become greater, but this 

 does not affect the velocities). Why then 

 the change of temperature? In my opin- 

 ion the reason is as follows: 



Consider the first molecules near the 

 opening into the vacuous vessel. Suppose 

 a very quick molecule enters into the vac- 

 uum ; it will retain its high velocity. Sup- 

 pose further a slow molecule enters into 

 the empty vessel, it will soon be overtaken 

 by a quick one and exchange velocities 

 (according to the law of impacts between 

 elastic bodies) with it. In brief, for a short 

 time there will be, in the empty vessel, only 

 quick molecules, and consequently in the 

 other vessel only slow ones. We have 

 here, as it were, a separation (fractionally) 

 into quick and slow molecules which causes 

 the observed changes of temperature. 



The Use of Potassium-ferric Chloride for 

 the Solution of Steel in making the De- 

 termination of Carhon: Geo. Wm. Sar- 

 gent. 



Ferric chloride reacts upon the steel 

 drillings according to the equation: Fe + 

 2FeCl3=3FeCl2. The reaction takes 

 place best as in the case of the copper salt, 

 in the presence of potassium chloride. To 

 prepare the salt, dissolve 267 grams of the 

 c. p. ferric chloride of commerce, which 

 contains about fifty per cent, of FeClj, and 

 130.7 grams of potassium chloride in a liter 

 of water. Two hundred cc. of this solution 

 will decompose a factor weight of drillings 

 in the same time as the double chloride of 

 copper and potassium. Free hydrochloric 



acid should not be in excess of the ratio 

 of 1 cc. to 225 cc. The amount of acid 

 should be just sufficient to prevent the 

 formation of any basic salt of iron. By 

 chlorinating the filtrate from the carbon 

 (see Jour. Am. Chem. Soc, 22, 210), 

 it may be used to decompose a second lot 

 of drillings. This may be repeated until 

 sufficient iron has accumulated to permit 

 of the addition of more potassium chloride 

 and a further dilution. The solution after 

 it has decomposed a weight of drillings re- 

 mains transparent, unlike the copper salt; 

 hence the point when the decomposition of 

 the drillings is complete is readily ascer- 

 tained. This fact, together with the ease 

 with which one can observe how the filter is 

 holding the carbon, as well as the economy 

 of the process, recommend it. (The article 

 will be published in the Journal of the 

 American Chemical Society.) 



Condensation of Chloral with the Nitrani- 

 lins: Alvin Sawyer Wheeler, and H. 

 R. Weller. 



Chloral readily reacts with the three 

 nitranilins with the elimination of water, 

 forming condensation products. If the 

 temperature is kept down, addition pro- 

 ducts are first formed. The condensation 

 products are beautiful crystalline yellow 

 bodies. We name them ' trichlorethyliden- 

 di-nitrophenamines. ' The melting points 

 are: o-body, 171° ; m-body, 212° ; p-body, 

 218°, uncorr. The para-body was prepared 

 in 1898 independently by Eibner and by 

 Baskerville. 



The Electrical Conductivity of TJnne and 

 its Relation to Chemical Composition: 

 J. H. Long. 



In this paper six complete analyses of 

 normal urines were given along with the 

 electrical conductivity. It was then shown 

 that nearly the same conductivities are- ob- 

 tainable by combining the inorganic salts 

 in the proportions as found by analysis, 



