354 Profs. Percy Frankland and Marshall Ward. 



99 and 132 ; total = 231 ; mean = 115 per c.c. After the same 

 time F 12 (unexposed) gave 1600 and 1920; total, 3520; mean, 1760. 



The flasks were meanwhile put back and exposed yet another 

 4^ hours to the afternoon sunshine i.e., from 12.30 to 5 P.M. of 

 which about 3-ijr hours counted as bright sunshine. 



Two plates from F 11 made at 5.30 gave 248 and 279 ; total, 527 ; 

 mean, 268 colonies as the number per c.c. ; while two plates made at 

 the same time from the unexposed flask (F 12) were so badly lique- 

 fied that, although we estimated 3300 per c.c. from one of them, we 

 regard the numbers as really higher. 



These flasks stood in the laboratory over-night at a temperature of 

 18 C., and were then exposed next day from 11.30 A.M. to 4.30 P.M.,. 

 about one and a half or two hours of the five being sunny. Then, at 

 5 P.M., fresh plates were prepared. 



Two plates from the exposed flask gave 640 and 3200 per c.c. re- 

 spectively ; total, 3840 ; mean, 1920 per c.c. The numbers are not 

 very good, as there is such a great difference between the two plates. 



Two plates from the unexposed flask gave 12,800 and 16,000 per 

 c.c. ; total, 28,800 ; mean, 14,400 per c.c., again bearing out the con- 

 clusion that the sun has powerful action on the exposed water. 



XVI. 



To my mind one of the most important discoveries elicited from these 

 plate cultures of Thames water was the obvious reduction of liquefac- 

 tion on the plates made from water exposed to light, and so struck was 

 I with the differences between these plates and those made with the 

 water not exposed that I made an independent investigation into the 

 matter by selecting a set of the most pronounced liquefying forms 

 from the water and examining their behaviour when exposed to light 

 side by side with that of non-exposed samples. 



I started with the commonest and most pronounced liquefying 

 form in the Thames water at the time. I refer to it throughout as 

 Colony ft in my notes, and write it shortly /?. I had noticed the 

 following facts concerning it during our experiments on the action of 

 light on the Thames water as collected it should be borne in mind 

 that I had already studied its characters and knew the form pretty 

 well. 



In the first place the plates as a whole liquefied very much more slowly 

 and less completely than those made from unexposed water. Secondly, 

 although it seemed at times as if this was because the form /3 had 

 been eliminated from the water, I suspected that a certain other form, 

 which liquefied less rapidly and developed much more slowly alto- 

 gether, was really the above-mentioned form /3 with feebler charac- 

 teristics. 



