BY R. GREIG-SMITH. 



145 



pended moss, which were incubated at 22" for varying periods. 



The extract was prepared in the usual way by filtration first 

 through paper, then through porcelain. One portion of the 

 extract was boiled for an hour under an inverted condenser. The 

 various portions, generally SOc.c, of the raw, boiled, and diluted 

 extracts, were each seeded with 1 c.c. of a suspension of Bac. 

 2)7-odi(jiosus, and incubated at 22° for 20 hours, when counts 

 were made by the plate-method. The numbers of bacteria in the 

 raw extracts were taken as 1 00, and those in the treated extracts 

 were calculated in terms of this. Fractions were omitted, and 

 numbers less than 1 were taken as 1. The actual number of the 

 water-control can be found by dividing the extract by the 

 extract/water ratio, for the latter was obtained by dividing the 

 extract-count by the water-count. 



Experiment i. 



In looking over the results of Experiment i., it is seen that 

 moss is not a good substance for determining toxin-23roduction. 

 When new, it is too nutritive, and when old, that is, when it had 

 been used and washed once or twice, it is too poor. The dilution- 

 curves of tests 3, 4, and 5 are almost horizontal lines, indicating 

 that the extract is of a nature similar to water. The results 

 obtained by boiling the extract in tests 3 and 5 cannot be ex- 



