52 Report S.A.A. Advancement of Science. 



with a good pigment-forming strain of a green, very motile organism, 

 closely akin to the B. fyocyaneiis. A number of trials showed that 

 this organism could be easily recovered, both from distilled and tap 

 water, even after some days, and as the presence of fluorescein did 

 not interfere with its recovery, the organism was suitable for the 

 purpose, and the following experiments were carried out : — 



A tube, similar to that previously described, was prepared with 

 a column of fine sand, 7.5cm. in length, and the sand, apparatus, 

 etc., having been foMnd free from the pigment-producing organism, 

 50 cc. of a 0.02% -solution of fluorescein was made up to 1,000 cc. with 

 distilled water, to which a small quantity of an agar growth of the 

 organism had been previously added, and incubated for 24 hours : 

 the solution was fed on to the sand as in previous experiments. 



Sixteen minutes elapsed before the first drop fell from the tube, 

 and the green organism was present from the beginning and all 

 through the experiment, whereas it required about an hour before (in 

 the 19th cc.) the fluorescein solution came through, apparently 

 unchanged. 



An experiment carried out similarly with the same material, but 

 with a column 32cm. long, gave the same results : 69 minutes elapsed 

 until the first drop fell from the narrow tube, and the green organism 

 was present at once, but the fluorescence only appeared after some 

 25CC. had passed. 



After complete draining, water was added, and the green 

 organism was found in the 7th cc. of the washing, showing that, 

 although the experiment had lasted some days, the sand column had 

 not become a filter for the organism. 



The same result was obtained with a column of sea-sand, 32cm. 

 lf)ng, and also with a 32cm. long column of a very sandy soil. 



A black loamy soil (from Potchefstroom) was next tried. Through 

 a column, 32 cm. long, the percolation was very slow, 6 hours and 

 43 minutes elapsing before the first drop fell. Contrary to expecta- 

 tion, however, the first drop showed a very faint fluorescence. The 

 faint coloration persisted for a considerable time without apparently 

 becoming more distinct; after about 20 cc. had passed through, the 

 intensity of fluorescence began to increase, and after about 50 cc. had 

 percolated, the intensity was about as great as in the original solution. 

 There ai)peared in this case to be a slight filtering action with respect 

 to the organism, for on "plating" the first cc, the pigment did 

 not show until considerably after 24 hours' incubation. This may. 

 however, have been due to the presence of some substance extracted 

 from the soil (humic acid ?), inhibiting the growth or pigmenting 

 power of the organism. The organism was similarly found in the 

 iTth cv. and in the 30th cc, showing on agar after 24 hours 

 incubation. 



These results, and particularly those noted in connexion with 

 adsorption, indicate that fluorescein must not be expected to appear 

 in a time proportional to the rate of flow, when the water has to- 



