( (i()4 ) 



This is confirmed (inalitatix elv hul the exact determination of the 

 lines H'l and K'l and conseqnenlly also of the ternarv entecticum 

 proved to be impossible. 



In this way no more was ol)taine(l tiian is indicated in Fig. 1, 

 namely the solidifieatioji lines Ad and OD for the mixtures of Cl^ 

 and S^CIo in which no compound had formed and besides these the 

 series of tiie first solidifying points AH, HEK and KD for the par- 

 tially combined liquids formed at 0^. 



These lines therefore correspond with the accentuated lines in 

 Fig. 2. The true meitijig ])()int of SCl^, without decomposition ot 

 the liquid (E F'ig. 2) is therefore not yet known. If SCl^ could have 

 been prepared in pure crystals, this might have pi-obably been 

 determined as it would presumably be situated below the tempera- 

 ture where the liquid couqtound shows perceptil)le deconq)ositi()n. 



This is rendered more proitable by the observation that notwith- 

 standing their minuteness the crystals of SCl^ when rapidly heated, 

 are quite permanent up to — 20^ 



Fig. 1 does not agree witii tig. 2 in an important point. In 

 the latter DK is placed above DG, in the tirst DK is found below 

 DG. DG represents the lowering of liic melting point of S.^C\^ by 

 added Cl^ when this remains unchanged: DK when a portion of it 

 combines with a j»art of the S.X'l^ to SCl^ according to thee(|uation 



S.,C1, -f 3 Ci., = 2 SOI,. 



If we now compare the total amount of tlie foreign molecules 

 which occur along with the S.^CM.^, for the same total amount of 

 Cl.^, these numbers are smallci- in the case of a partial formation 

 of SCI4 than when this is not the case, xlccording to this suj)position 

 DK ouglit to be situated liigher tlian DG. 



As in practice, the reverse position was found Ave must look for 

 a cause which may exi)lain this fact. AVc find this by sui)posing 

 that besides SCI, tiiere is also formed in tiio hipiid a considerable 

 proportion of SCl^ accordiuu' to the e(|ua1iou 



S.,C1., 4- Gl, = 2 SGI,. 



By taking this view the ]iumi)er of foreign molecules mixed 

 with the SjGl., becomes greater ow ing to the formation of SG1._, than 

 without this. 



Tlie formation in the liquid of an amount of SGlg exceeding that 

 of SGI, \vhicli had already become prol)able by the dilatometric 

 experiments has, therefore, been confirmed. The slow crystallisation 

 of SGI4 will now be better understood; but the liquids between Gl., 

 and SjGlj should now be considered not as ternary but as quater- 



