( 468 ) 



tiglitlv rulled piece of platinum ,uauze in llie sul[)lnir; il had aliuosi 

 the shai)e of a nail, and rested with its head on the bottom of the 

 tube. Our purpose wtis to diminish tiie differences of temperature in 

 the cooling sulplmr in this way, and really under these circumstances 

 there was jio question of the appearance of a second layer, at least 

 decidedly not in the immediate neighbourhood of the metal ; at greater 

 distances, however, streaks were now and then observed, as was to 

 l)e expected. 



Further the experiment was made with capillaries. A current of 

 liquid along the wall was out of the question here, and so it was 

 found that the |)henomen(»n of the layers occui'red neither in case 

 of rapid cooling down from 450° in a paratHin bath of 100°, nor in 

 case of cooling in the air (with or without blowing). We could 

 however see the liquid assume a lighter colour tirst at the bottom 

 as was to be expected, l>ut the transition from light to dark was 

 always continuous here. 



In the experiments mentioned iqj it) now the sulphur was continually 

 in contact with S().,-containing air, and as it stands to reason that 

 the application of a positive catalyser according to our views will 

 rather intensify than prevent the two-layer phenomenon, some experi- 

 ments were made with sulphur catalysed by NH,. (|)uite in accordance 

 with what was expected it a|>peared that the two-layer idienomenon 

 manifests itself in this case in such a splendid way that we can 

 urgently recommend the blowing through of NH,. during the [)eriod 

 of cooling for a demonslral ion-experiment. 



In conclusion it may be poiiUcd out that Wkjam)') has found that 

 in agi-eement \vilh the regularity found by Smits and Aten") the 

 ecpiilibriiini in the liipiid sulphui- is shifted to the endothermic side, 

 i.e. to greater S„-concenlralions by illumination. H<»\v great this 

 shifting iS at different temperatures iias not been stated as yet, but 

 it has been ascertained at this laboratory, that there is no question 

 of unmixing then. 



liniKti-k. If it appears in further investigations as is to be expected 

 that as Smith and Carson think (Zeitscln-. f. phys. Chem. 77, ötU 

 (19Ji)), a third crystallised moditication is to betaken into account, 

 figs. 1 and 2 will have to be modified in accordance with this. 

 Anon/ffiiic Clwrnicdl L(il>or(tforiiy of the {/invei:siti/. 



Anistenbtiii, October 1911. 



1) Zeilschp. f. phys. Chem. 77, 4:23 (i'Jfl). 

 ■^) These Proc. Oclobei' 30 (l'.)Ü9i. 



(November 22, 1911). 



