488 



Profs. Liveing and Dewar. 



[Dec. 11, 



temperature and the action of sodium, wherewith to close the lower end 

 of our tube of inspection. We have tried a plate of colourless 

 sapphire, but it very soon became too opaque for further use. Perhaps 

 i'used alumina may do better. 



A few experiments made in a similar way with thallium in iron 

 bottles, showed on the whole similar results. No bright green line was 

 seen on dropping in a piece of thallium, but it was not until after the 

 lapse of some seconds that the absorption line was apparent. It was 

 faint at first, gradually deepening, and soon attaining a maximum of 

 intensity, when it was remarkably stationary for a long time. More 

 pieces of thallium introduced successively produced a small but per- 

 ceptible widening of the absorption band and a diminution of the 

 sharpness of its edges. 



The foregoing account gives the general course of the phenomena, 

 and it seems useless at present to describe further details, because the 

 quantitative results which we hoped to give are for the time in abey- 

 ance until we can ascertain more correctly the temperature of our 

 furnace and the thickness of the layer of vapour observed. This 

 requires an accurate knowledge of the density of the vapour, which 

 could not be satisfactorily deduced from the experiments made with 

 iron bottles, as they did not accord well and were complicated by the 

 action of iron on sodium above described. The platinum vessels stand 

 but few experiments without repair, which causes considerable delay ; 

 and the above-mentioned results have necessitated the expensive 

 operation of the reconstruction of some five platinum bottles. 



The phenomena attending the compression of the vapours, as well as 

 those of the amalgams of varying percentages, seem to indicate that 

 the width of the D absorption is dependent on the thickness and tem- 

 perature of the absorbent vapour rather than on the whole quantity of 

 sodium present in it. Yery minute quantities diffused into the cool 

 part of the tube appear to give a broad diffuse absorption, while a 

 layer of denser vapour of small thickness in the hottest part of the 

 vessel gives but a very narrow absorption. This may, however, be 

 due to the variation of temperature. 



The phenomena of the bright D ban^. seen on compressing the 

 vapour is due to the elevation of temperature consequent on com- 

 pression. This elevation, we have calculated, for a sudden com- 

 pression to half the original volume, will amount in the case of sodium 

 vapour, if the ratio of the specific heat at constant pressure to that at 

 constant volume be taken as 1*14, to about one-third of the original 

 temperature, whatever that temperature be ; so that, allowing for loss 

 during slower compression, the temperature of the vapour in our 

 bottles will have nearly reached 2,000°. In the case of the amalgams 

 the main portion o^ the gas being mercury vapour, the effect of com- 

 pression to a like amount will increase the temperature by nearly two- 



