ENDOMIXIS IN PARAMAECIUM AURELIA 49 



In the present study I have made a sufficient comparison of 

 the osmotic pressures of several cultures used in isolation experi- 

 ments to show that the differences are probably too slight to 

 explain the high death rate which is found in many cases. In 

 order to ascertain whether there is any close numerical corre- 

 spondence between the percentage of loss in such isolation ex- 

 periments and the percentage of multinucleate specimens in the 

 culture at the time of isolation seven experiments were per- 

 formed. In each of these the Paramaecia were isolated with a 

 capillary pipette, and placed on cover glasses on slides in a 

 small slide stack, which was kept in a moist chamber. The exact 

 amount of isolation fluid was not measured. It was, however, 

 in every case many thousand times the volume of one Paramae- 

 cium and sufficient to maintain many individuals. The kind of 

 culture fluid used varied in the different experiments In ex- 

 periment 18 an old hay infusion in which Paramaecia had been 

 living for some time was employed. In experiment 19 an infu- 

 sion of grass and in 20 and 21 a fresh infusion of hay was used. 

 In experiment 23 and part of 22 a sterilized 0.025 per cent solu- 

 tion of Liebig's extract of beef in distilled water was used and in 

 24 and part of 22 the parent culture (hay infusion) The num- 

 ber of Paramaecia which died within about forty-eight hours was 

 noted, which number gave the percentage death rate A num- 

 ber of individuals chosen at random from the culture to be tested 

 were stained and mounted, from which the percentage of multi- 

 nucleate specimens was determined. The number so chosen 

 varied in different experiments but was sufficiently large in 

 each to give a fair estimate of the latter figure. Further details 

 of these experiments may be omitted, the results being sum- 

 marized in the following table. 



In only two (18 and 19) of these seven experiments is there 

 any correlation between the death rate and the percentage of 

 multinuclearity at the time of experiment and this is probably 

 accidental. The objection may be brought against these experi- 

 ments that different cultures were employed in different cases 

 and therefore comparisons are misleading. What I believe I 

 have shown, however, is that there is no relation between death 



