( «48 ) 



ai leiiiiK'ialui'Os of 103 — 110 ; llic resulls ohlaiiied will be coniniu- 

 nicated on a fiitnre occasion. 



3. It was obvious that 1 should trv to make a closer study of 

 tlie velocity of ti-ausfonnatiou of the iodide^ in solution. Its great 

 insolubility, however, rendered the operation ini|tossible: of the many 

 solvents \vhich were tried, pyridine ju'oved to be the best ; tiie 

 solubilit}' of the iodide Avas however still too smnll, namely only 

 abont l\/.//„. 



The more soluble dijtlienyliodonium-cldoride was better suited for 

 the pnrposc; the solubility in water, althouuii not large, proved suHi- 

 cient at the temperatni-e at wiiich the opei-at ion took place (98 — iHr). 



The results obtained in lifleen \elocity determinations were at 

 lirst very unsatisfactory and pointed to the existence of many inter- 

 fering influences. The coefiicionts obtaincMl on applying the formidae 

 for niuniolccnlar and bimolecnlai- reactions were anything l)ut c(mstant 

 and often })ointed to a very irregular course. In one expei-iment 

 coeflicients were obtained which were many linu^s gi-ealer than those 

 got in another ap})ai'ently (piite analogous case. Sometimes thecoefti- 

 cients diminishe<l ecpially. sometimes the i-eaction after proceeding 

 for a while, suddenly ceased. After many similar negntixc results it 

 at last appeared that the conxci-sion of diplieny liodouium-chloride into 

 cliloro- and i(»doi>eu/.en(^ is inlbienced lo au extrnordinary degi-ee by 

 the presence of \ery small (piantilies of impui-ilies. \ ei"y small 

 quantities of acid retard the r(Mction to a remarkal>le extent or bring 

 it lo a standstill: the presence of traces of iodine causes n regular 

 fall in the reaction coeflici(Mit ; a little of the tree base (dipheuyl- 

 iodonium-hydroxide) accelerates, ou the otliei' hand, the decomposi- 

 tion in a strong degree. The halogeid>en/enes foruied during the 

 reaction appeared hoivever to be inei-t. 



( )n now using a \ery pui-e pre|)aralion free from acidity and of 

 a pure white colour and applying the forunda for i-eactiojis of the 

 second order, ct)eflicients were ol)taiued w liicli couhl be considered 

 as constants, (see table p. (545)). 



On adding (i cc. of "Z.^,. fK-l, the coetïicient (which, moreover, was 

 not constant) fell to about half the \alue obtained in experiuieni 1 

 while the presence of 8 cc. of "Z^. ((.'..ilj.^ J( )11 increased the coeflicieut 

 abont 4 or 5 times ^). 



1) Tlio following expoi-iment also may sliow how spiisilivo llie transfoniialioii is 

 to very trifling qiianlitios of foreign sulislancos. 'fo a solnlion of tlio cliloridt» 

 {^'a~\'-2i^-2, T — O'l.O), wlii.li after V/., lioiirs had fallen from ;50.(i7 AgNO.. to 

 23.71, was added 3'..>.3 iiiilligr. of a well crystallised cliloriile wliieh was coloured 



