82 THE GROWTH AND DEATH OF BACTERIA 



of these a number of tubes of medium, usually liquid, are inoculated with suitable unit 

 quantities. The tubes are incubated and from the number in which growth occurs, the 

 probable number of bacteria in the original suspension may be calculated from formulae, 

 such as those worked out by McCrady (1915), Stein (1917), Greenwood and Yule (1917), 

 Halvorson and Ziegler (1933a, h), and Gordon and ZoBell (1938). (For useful tables see 

 McCrady 1918, Hoskins 1934.) This method, or one of its numerous modifications, is 

 frequently adopted when an approximate estimate of the numbers of living bacteria in a 

 suspension is required. It has the great advantage of being applicable to organisms that 

 cannot be counted by the ordinary plating method. It is used, for example, in the deter- 

 mination of coUform organisms in water, because no satisfactory plating medium has yet 

 been devised for differentiating coUform bacilli in mixed culture from other bacteria, though 

 several suitable liquid media are available. It is frequently used in filtrable virus work ; 

 for this purpose quantities of the dilutions are inoculated into a susceptible animal, generally 

 by the intradermal route, and the occurrence of a specific skin reaction is regarded as 

 evidence of the presence of the organism in the corresponding dilution. As Halvorson 

 and Ziegler (1933a, h) have pointed out, the method is subject to a very large experimental 

 error, depending mainly on the numbers of tubes seeded from the different dilutions. Even 

 with 40 tubes to each dilution the count is liable to vary between about 38 per cent, below 

 and 47 per cent, above the true count, while with only 5 tubes to each dilution the cor- 

 responding figures are — 70 and + 260 per cent. The method is not therefore suitable for 

 exact bacterial enumeration. 



(6) The Plating Method. — This is generally performed by a modification of Koch's 

 original platmg method. It consists essentially in preUmmary dilution, if necessary, of 

 the suspension, the plating out of unit quantities of suitable dilutions into a suitable 

 sohd medium, and the counting of the number of colorues that develop after incubation. 

 The average number of colonies per plate multiphed by the reciprocal of the dilution 

 affords an estimate of the number of living organisms in the original suspension. Instead 

 of plates, roll-tubes may be used. Provided that the bacteria are homogeneously dis- 

 tributed in the suspension, that not more than one species of organism is present, and that 

 attention is paid to a large number of technical points, accurate counts may be obtained 

 by this method. Departure from these provisos may often entail, however, experimental 

 errors of considerable magnitude. (For sources and measurement of errors see Wilson 

 et al. 1935, Jennison and Wadsworth 1940.) 



Miles and Misra (1938) have described a method of plate counting in which the bacterial 

 suspension, instead of being mixed with the melted agar, is deposited in the form of drops 

 on the surface of the sohd medium and the count estimated from the number of colonies 

 that develop. 



The Growth Curve 



If a given bacterium is seeded into a liquid medium of suitable composition ; 

 and incubated at a suitable temperature, it will be found that its growth will follow 

 a definite course. This course is most conveniently represented in graphical form 

 (Fig. 15), the logarithms of the numbers of bacteria along the ordinates being 

 plotted against the time in hours along the abscissae. 



The growth curve may be arbitrarily divided into four phases : (1) The lag 

 phase, a to h, lasting for a few hours, during which multiplication is slow. In the 

 early part of this phase there may be no apparent growth ; in fact many of the 

 organisms may die, so that there is an actual diminution in their numbers. Within 

 a short time, however, growth becomes apparent, and gradually increases in pace till 

 the beginning of the next phase. (2) The logarithmic phase, h to c, in which regular 

 division of the organisms occurs at maximum speed. Since their increase is in 

 geometric progression, it follows that when the logarithms of their numbers are 

 plotted against the time in hours they fall on an ascending straight Hne. (3) The 



