178 Appendices to Foihrth Annual Report 



decomposition has been commenced by bacteria, and where such decom- 

 position has gone on for a long time. The presence of bacteria in waters 

 is, then, a most important factor in connection with the presence of fungi 

 in our fish streams. Firstly, as an index of the amount of organic matter 

 present, and secondly, in connection with the action upon this organic 

 material, before it can contribute readily to the nutrition of fungi. 



How far this fact may have a potent influence on the presence and 

 persistency of the salmon fungus (Saprolegnia) in some rivers, must be a 

 subject of future investigation. In the water from the Tweed, which 

 contained the largest number of bacteria, the salmon disease was preva- 

 lent in abundance. 



For the purpose of determining the number of organisms present in a 

 sample of water, Koch's plate gelatine method was employed. This has 

 already been partly described."^ It is based upon the fact that most 

 common micro-organisms grow readily on a gelatinised broth, to which a 

 certain quantity of dried peptone has been added. A definite quantity of 

 water is mixed with a definite quantity of this gelatinised broth, pre- 

 viously liquefied by warmth, and the whole spread upon a glass plate in a 

 thin uniform layer. The layer of jelly is kept under cover at a suitable 

 temperature for from two to six days. The individual bacteria or other 

 organisms are widely enough separated to ensure separate germination, 

 and from each there develops a separate colony, which soon becomes 

 sufficiently large to be seen with the naked eye. 



In order to count the number of colonies, which corresponds, of course, 

 to the number of bacteria or spores of other organisms present, a glass 

 plate, upon which has been ruled with a diamond a number of lines 

 dividing it into equal squares, is laid over the layer of jelly, and the number 

 of colonies counted in each one of a number of squares. The total number 

 of colonies, when divided by the number of squares counted, and multi- 

 plied by the number of squares which correspond to the total area 

 of the jelly, gives the number in the quantity of water examined. We 

 have, however, adopted a still more exact method, viz., to number the 

 squares, and to count every one of them separately. In this way an 

 absolutely exact result is obtained, and any error can be checked by 

 counting the same square over again. 



Th,e number of organisms is usually calculated by taking as the unit one 

 gramme of water ( = 1 cubic centimetre at 18° c). It is well, however, to 

 use a smaller quantity than one cubic centimetre, and to determine the 

 exact quantity by weight rather than measure. This is done by using 

 uniformly a pipette of definite size, graduated in the following manner : — • 



A piece of filter paper was carefully weighed, and then five drops of 

 water dropped upon it from the pipette to be used, and the whole 

 weighed again. Five drops more were added, and the whole again 

 weighed. After numerous repetitions of this process, it was found that 

 with a uniform method the weight of five drops was so constant as to be 

 perfectly reliable. I»n the case of the pipette used the weight of five drops 

 was '27 gramme (and the number of organisms per gramme was, of course, 

 easily arrived at by multiplying by 3'7). 



Although at first sight this method may appear rough, it has the great 

 advantage of avoiding risks of other contamination of the water, 



A more strictly accurate method is that adopted by Dr AVinter Blyth, 

 who takes a quantity of the water to be examined in a small drop bottle 

 previously sterilised. The water and bottle are weighed, then a few 



* For the exact details and an account of the precautions taken to ensure sterilisa- 

 tion and the exclusion of germs from the air, see Third Annual Report of Fishery 

 Board for Scotland, 1885, p. 75. 



