344 Profs. Percy Frankland and Marshall Ward. 



After another day's incubation one plate was liquefied; the others 

 gave 832, ]920, and 816. Total of three plates = 3568 ; average = 

 1189 colonies per c.c., and this may be taken as the highest number 

 obtainable, the counting having been done and checked very carefully 

 and thoroughly with a good lens. 



After the five hours' exposure, new plates were made two from 

 each of the flasks and, in order to obviate as far as possible our 

 previous difficulties with the liquefying forms, we diluted each 1 c.c. 

 of the sample water with 9 c.c. of sterile distilled water, carefully 

 prepared in advance. 



It is unnecessary to give details as to sterilisation, &c., but the 

 following are the essential points of the plan followed : For each 

 sample to be taken two pipettes and two test-tubes are needed. One 

 of the test-tubes of each pair is graduated to hold 9 c.c. of the sterile 

 water; into the other 1 c.c. of the water to be tested is dropped with 

 one pipette, and the 9 c.c. of sterile water are then poured on to this, 

 thus ensuring thorough and rapid mixture. The second pipette is 

 then used to obtain the one or more drops taken to make the gelatine 

 plate. 



By this method we found that two plates from F 3 (the exposed 

 flask) gave 2880 and 1280 per c.c. on the 4th day. Total of two 

 plates, 4160; mean, 2080 per c.c., suggesting that the light was not 

 strong enough to prevent the bacteria from multiplying. On stand- 

 ing two days longer these plates gave 3840 and 1600 ; total, 5440 ; 

 mean, 2720 colonies per c.c., showing that there were a good many 

 slowly developing germs present, and we were struck with the paucity 

 of liquefying forms. 



Two plates from F 4 (the unexposed flask), made and examined at 

 corresponding times, and in the same way, gave 660 and 2310 per c.c. 

 on the fourth day. Total, 2970 ; mean, 1485 colonies per c.c. ; and 

 on the fifth day (we could not go further) 660 and 3630 ; total, 

 4290 ; mean, 2145 colonies per c.c. 



So far, therefore, the numbers in the two flasks did not appear to 

 be appreciably affected by the little sunlight that reached the water. 



The two flasks were now placed in an ice-safe over night, for expo- 

 sure next day. They remained on ice about 14 hours. 



Next morning, August 23, these flasks were again put out at 9 A.M., 

 and remained till about 4 P.M. (over seven hours) ; but it rained 

 steadily all the time except the last hour, when the sun shone. 



Two plates, made of diluted samples as before, from each flask, 

 were made at noon on this day, with the following results : After two 

 days' incubation the plates from flask F 3 (exposed) showed 39,600 

 and 37,620 per c.c. Total of two plates, 77,220 ; mean, 38,610. After 

 a further day's incubation the plates gave 47,190 and 41 250 ; total, 

 88,770 ; mean, 44,385 per c.c. 



