SEA-FISHERIES LABORATORY. 161 



the tangent which the line makes with the axis of x, and a the 

 intercept on the centroid vertical of the graph. Remembering 

 that these are really logarithms we now find the smoothed 

 values of the original data. 



We also find the equation to the hyperbola. Converting 

 the logs back again into natural numbers we find the value 

 of n, now the index of x and positive ; a is, of course, merely 

 a scale constant. It represents the (graduated) number of 

 bacteria per unit volume which was present in the culture at 

 the beginning of the experiment. We find the constants in 

 the equation yx n = a. 



III. The Counts of Colonies on Plates. 



Some matters of interest with regard to methods may 

 be noticed here. As a rule one finds different kinds of colonies 

 on the same plate. In cultures of badly polluted or recently 

 polluted mussels in neutral red, bile-salt, lactose agar we 

 find both crimson and colourless colonies. The latter are 

 always on the surface of the plates ; at least, if they are present 

 in the depth of the medium one cannot see them. They 

 differ greatly in their ability to ferment sugars from the 

 crimson colonies ; as a rule they do not ferment glucose and 

 lactose. They are hardly ever seen in cultures from mussels 

 which are not badly polluted, and which have been taken 

 from places at some considerable distance from the source 

 of pollution. They are always absent in cultures made from 

 mussels which have been re-laid. I attach some importance 

 to the presence of such colourless colonies as indicating recent 

 pollution. I have subcultured very few of them — none of 

 these subcultures are described in this Report. As a matter 

 of both theoretical and practical interest they ought to be 

 closely investigated. 



The crimson colonies are present both on the surface 



