1^1-J Goodspeed: Xicotiana Hybrids 139 



upon which dates the flowers were measured between 5 and 6 

 p.m. The greatest effort was made to exclude personal bias in 

 connection with the measurements. As the plants were worked 

 with nearly every other day for over three months it was im- 

 possible not to form a more or less definite idea of the direction 

 towards which the results were tending. Also, when measuring 

 fifty to sixty flowers on one day from a single group of plants, 

 one could scarcely avoid keeping certain limits of fluctuation 

 in mind. It was attempted, however, to make the measuring 

 as mechanical an effort as possible and the actual limits of 

 fluctuation, as given below, were not found until practically 

 the last record was taken. The averages of the measurements, 

 also, were not calculated until all the measurements were com- 

 pleted and the experimental portion of the work concluded. 



5. Results of the Measurements 



The following table gives the number of days upon which 

 the flowers in each group were measured, the largest and smallest 

 number of flowers measured on a single day, the total number 

 of flowers measured, the largest and smallest diameters and the 

 average diameter in each case, the theoretical size of the true 

 intermediates between the average diameters of the parents, and 

 the fluctuations in corolla diameter throughout. 



Group number Var. I II III IIIJx I<;;J' IJX IH^ I$X IIJ' HJXIIIJ' III$X H;;? 



Number of days upon 

 which plants were 

 measured 



Largest number meas- 

 ured on a single day 



Smallest number meas- 

 ured on a single day 



Total number measured 



Diameter of largest 



flower, mm. 29 22 15 27 23 30 20 



Diameter of smallest 

 flower, mm. 



Average diameter, mm. 



Intermediates between 

 averages of parents, 

 mm. 



Fluctuation in corolla 

 diameter, mm. 



