Report on the Bacteriology of Water. 429 



labour involved would be too great to charge myself with at present, 

 sufficient facts came to hand to show that this branch of the enquiry 

 also ought to be taken up and pursued. 



Fourthly, the point arose by the way how far the doubling period 

 at a given temperature, and in a given medium, is affected by 

 whether the filament measured has developed all along, i.e., had 

 germinated at the temperature of the growth measurements, or 

 whether germination was conducted at some other temperature, and 

 the growing filament then brought into the new temperature. 



A fifth point arises by the way also. How far is the matter 

 affected by the age of the food-medium ? 



Curves of the Doubling Periods. 



If we plot out the results reached so far, some interesting curves 

 can be obtained which help us materially. 



Taking any series say the cultures in weak gelatine the averages 

 of all the doubling periods obtained for the several temperatures 

 may be plotted on ordinates erected on a base line divided into 

 degrees centigrade, and on joining the points thus plotted we get a 

 curve convex to the base line. As a matter of fact, however, the curve 

 so obtained from averages of all doubling periods is very irregular, 

 and all that one can say is that, (1) generally speaking, the curve 

 descends from somewhat high numbers below 18 C., to the lowest 

 numbers between 25 30, and then ascends again. 



2. The various jerks on this curve are unequal, and without any 

 evident order, and the re-ascent of the curve beyond 30 C. is not 

 always obtained. 



3. On the whole the curve descends most for broth, and least for 

 weak gelatine. 



It seems hardly worth while to go more into details, since it was 

 so evident that something had to be explained here, before further 

 steps could be taken, that I made a careful analysis of all the condi- 

 tions, and found that the irregularities on these curves of all averages 

 were due to a number of more or less controllable causes. 



In the first place, it turns out that it was a serious error on my 

 part to start the germination of the spores in the Sachs' box at 

 various convenient temperatures with reference to the temperatures 

 of observation, and many of the irregularities in question are due to 

 the consequent (1) differences in initial vigour of the plantlets 

 started at different temperatures, (2) differences in phase of growth 

 of such rodlets, and (3) differences in degree of exhaustion of the 

 food-material in the drops, which had thus been exposed for different 

 lengths of time to different initial temperatures. 



It is hardly necessary to go into the details of this very complex 



