R. A. Fisher, H. G. Thornton, and W. A. Mackenzie 345 



variation between parallel plates, for unless present in very large numbers 

 it would not appear on all the plates or even in every batch of five plates. 



An example of high variation between parallel plates, that was 

 actually traced to such an organism, is given to illustrate this cause of 

 inaccuracy. 



The soil used in this case was from the Leeds Experimental Farm, 

 and had received a treatment of naphthalene. Thirty parallel platings of 

 this soil were made on Thornton's agar. The counts of colonies on these 

 plates are given in Table XlTl. 



Table XIII 

 Parallel plates of Leeds soil 



It will be seen that the variation between parallel plates in the whole 

 series is excessive. In examining the plates, some were found to contain 

 an organism forming a growth between the agar and the bottom of the 

 dish. This organism occurred on the plates italicised in Table XIII. 

 It is a motile organism and apparently spreads in the water film under- 

 lying the agar. On plates 28, 29 and 30, the growth of this organism 

 was sheet-hke and from the low counts obtained it would appear that 

 its growth has reduced colony development. On plates 1, 2, 3, 4 and 6, 

 it has produced a number of separate colonies underlying the agar. 

 These colonies were probably produced by individuals which had multi- 

 phed and migrated along the bottom of the dish after the agar had set. 



