212 BOTANICAL GAZETTE [OCTOBER 
zygospores would be expected does not alter the sex of the races in- 
~ volved. When they take part in any sexual reactions at all, they are 
consistently either plus or minus. The first reaction of the sexual 
process is sufficient to indicate their sex, and gives an index of their 
sexual vigor. In calculating the mean grades of sexual activity for 
the different races, therefore, it has been considered fairest to give 
the imperfect reactions equal weight with zygospore formation. 
What are the causes which prevent oné combination in a quad- 
rangular reaction from carrying the sexual process through to 
completion is a question requiring further study. In certain cases, 
465 plus H 457 minus 
Z O Z 
459 minus 4 217 plus 
at least, the distance between the inoculations of opposing strains 
seems to be a matter of some importance. In Circinella spinosa 
it has always been necessary to inoculate the opposite sexes very 
close together in order to obtain zygospores, which are not formed 
beyond a few millimeters from the points of inoculation. In a few 
cases a retest of an imperfect reaction between races of C. berthol- 
letiae, but with inoculations close together, has shown zygospore 
formation. . 
Certain contrasts which were repeated with inoculations at the 
usual distance apart gave different reactions from those first 
obtained, as may be seen by a comparison of table VII A with 
table VIII where the retest contrasts are listed. In table VIII 
there is, as might be expected, a certain amount of change in 
the grades assigned to the strength of the reactions. The present 
