14 KANSAS UNIVERSITY SCIENCE BULLETIN. 



In both of these the irregular character of the new race was 

 reproduced in the offspring, though the irregularity was 

 scarcely greater than that seen in the offspring of check 

 spores. 



A similar tendency towards elongation in cultures from old 

 spores was observed by Hansen in Saccharoymces ludivigii. 



As regards the formation of pellicles, there seems to be no 

 constant difference between the new races and the check. Pelli- 

 cles are quickly formed in both, and the cells composing them 

 show the morphological differences characteristic of the two 

 types. 



The stability of the new races has been tested not only by 

 three years' cultivation in various media, but also by two 

 series of selection experiments. 



In one series an attempt was made to determine whether a 

 new race could be made more filamentous by the selection of 

 the more elongated elements. A single, much elongated cell, 

 usually at least five times as long as broad, or a group of 

 united cells including one or more of such filaments, was 

 isolated, and from its offspring a similar selection made. 

 Eleven such selections were successively made, for the most 

 part in hanging drops of glucose bouillon. It was found 

 that during the experiment there was a greater tendency to a 

 more mycelium-like form of cells, but unless kept up by con- 

 tinuous selection, the filamentous type reverted to the orig- 

 inal form of the new race. So there was no evidence that 

 the type could be permanently changed in this direction. 



In another series selection was made in an opposite direc- 

 tion. Among the elongated cells of the new races there is 

 almost always found a proportion of spherical cells, not 

 grouping or otherwise visibly differing from the parent type. 

 From a new race, taken about twelve days after its origin, 

 and exhibiting well-marked race peculiarities, selection was 

 made of a single spherical cell, or of a spherical cell with its 

 attached bud, and from the offspring of this cell a similar 

 selection was made. A check of unselected race cells was 

 carried on under parallel conditions. While there was some 

 variation in the degree of sphericity of cells, the check 

 showed the same changes ; and at the end of a series of ten 



