38 Investigations on the Life-History 



as possible the same amount of material was, in every case, taken by 

 means of a straight needle. Second roll-tubes were inoculated from these 

 in almost every case. 



The colonies which appeared in the tubes were counted day by day after 

 Inoculation until their number became stationary, or so large as to be 

 countless, or until liquefaction of the gelatine prevented any further 

 enumeration. When rapidly-growing and liquefying colonies rendered 

 the gelatine fluid and inhibited the growth of the slower non-liquefying 

 forms, the number of the latter was estimated from their colonies 

 which began to grow before the liquefying growths had obscured 

 their further development. 



When the colonies were too numerous to count they were noted 

 as being " in large numbers ; " when still more numerous, and covering 

 the whole surface of the gelatine with minute growths, they were termed 

 " innumerable." When it was wished to compare the relative numbers 

 cultivated from different classes of fish, actual figures had to be substi- 

 tuted for these terms. In each case 250 has been used to represent " a 

 large number," and 500 to represent " innumerable." The actual figures 

 were certainly higher than those chosen, but those used indicate a 

 somewhat definite proportion between the actual numbers present, and 

 their constant substitution introduces no important element of error. 



In this part of the research valuable help was rendered by Mr. Hume 

 Patterson, the Laboratory Assistant. 



I. NUMBER OF MICRO-ORGANISMS. 

 (a) The Salmon Caught at the Estuaries. 



The chief details of the results obtained from the bacteriological in- 

 vestigation into the salmon caught in tidal waters are shown in Tables 

 II. and III. The results are grouped according to the season of the 

 year at which the observations were made. The number of the fish 

 corresponding to that in the general Laboratory register is first given, 

 and the year and month of capture are then noted: 



[TABLES. 



