D. Ward Cutler and L. M. Crump 17 



Soil, which had been sterilised in the autoclave under 15 lbs. pressure 

 for half-an-hour, was inoculated witli a counted suspension of trophic 

 and cystic protozoa. It was then treated with 2 percent. HC1. overnight, 

 and the numbers counted by the dilution method. Fifteen such experi- 

 ments were performed, given in Table II. On five other occasions two 

 samples have been taken near to each other from the same field, and the 

 number of protozoa in each sample found independently by two ob- 

 servers. The results so obtained were in every case identical. 



Uneven distribution over the field. This possible source of error appears 

 to us to have been obviated by the method of sampling we adopted. 

 Reference to the map on p. 13 will show that if uneven distribution 

 over the field is to account for the periodicity one must assume that 

 active and cystic flagellates occupy alternate situations over the 14 rows 

 which we sampled. Surely an impossible assumption ! 



The following experiments, however, will afford proof that the 

 protozoa are fairly evenly distributed. In 1916-17 one of us took soil 

 samples indiscriminately over the plot. These samples were counted 

 and the results are given below. 



On March 9th, 1920 (two days after the daily sampling had ceased) 

 two further samples were taken, B and C, and counted independently. 

 The numbers were as follows : 



Total Active Cystic 



Flagellates Amoebae Flagellates Amoebae Flagellates Amoebae 

 Series B ... 9,000 2,500 6,500 2,150 2,500 350 



„ C ... 9,000 3,250 5,750 2,650 3,250 600 



These figures justify the belief that unequal distribution in the field 

 cannot account for the flagellate periodicity. 



Death of cysts due to action of 2 per cent. HC1. In the previous publica- 

 tions referred to above (5, 6) it was shown that a small proportion of the 

 cysts might be killed by the acid; the proportion, however, was on the 

 average below 15 per cent., which could not account for the marked 

 differences obtained in the present series of experiments. 



Ann. Biol, vn 2 



