BARBER: HEREDITY IN CERTAIN MICRO-ORGANISMS. 35 



more than H, While H shows little microscopical difference 

 from its type, Y, like A, tends to form flocculi in bouillon, 

 H has less tendency to form very long filaments than either 

 A or Y, but the average length is much greater than that of 

 the type. I have at present all three of these new races grow- 

 ing in my laboratory, each with a check taken from the type 

 culture at the time when the new race began. 



Another sort of new race, characterized by a nearly com- 

 plete loss of motility, was obtained from an agar culture of the 

 same B. coli, type I, from which new races A and H had come. 

 In the water of condensation from this agar culture a series 

 of single cells were isolated, each in a separate droplet. A 

 number of these grew into normal motile cultures ; but one 

 of them, which at the time of isolation had the appearance of 

 a normal cell, gave rise to offspring almost wholly non-motile. 

 These did not differ in any other respect from type cells, ex- 

 cept that there was a tendency to form groups of short chains 

 lying parallel to one another ; and the tendency to adhere in 

 short filaments, seen in all very young cultures, persisted 

 longer in this new race ; both tendencies are perhaps due to 

 the loss of motility. 



In a large hanging drop from a fresh culture of this non- 

 motile race a few motile cells could usually be seen, and it 

 was found that repeated selections of these motile cells pro- 

 duced cultures of somewhat increased motility. On the other 

 hand, six successive selections of non-motile cells from the 

 original non-motile race has given a type which remains prac- 

 tically non-motile after a period of cultivation extending over 

 nineteen months. Over a year after its origin selections from 

 the very few non-motile cells appearing in cultures of this 

 race have failed to produce a motile type. 



Tests have been made of the vigor of growth and the power 

 of fermentation of this non-motile race, and it apparently 

 equals the parent type in these respects. An attempt to re- 

 store its motility by repeated transfers of fresh cultures grown 

 under favorable conditions of medium and temperature gave 

 negative results, though seven such transfers were made 

 during a period of three days. Similar transfers of the pa- 



