138 University of California Puhlications in Botany [Vol. 5 



in 1910 and also among themselves except as follows. One plant 

 resulting from the cross I$X HI J* was in habit exactly as all 

 the plants of variety I in 1911 and as all but one in 1910. Three 

 plants of I 5X He? were not up to the normal in height and 

 were more broadly spreading, with thicker main stems and lat- 

 erals. Lastly, five plants of IIIJX II c^ were very small and 

 in appearance immature, with few small laterals and small leaves. 

 The traces of fasciation observed in two plants of variety III in 

 1910 appeared in five plants in 1911 but only in those hybrids 

 which had variety III as a male or female parent. 



4. IMethod of Measuring Corolla Dlvmeter 



Measurements of corolla diameter of the parent and hybrid 

 flowers, were first made on September 8 and the last record 

 noted is on November 20. Slightly over 2750 measurements were 

 recorded during this period. A sheet of zinc. 3 by 4 inches, was 

 deeply notched and served as a very simple holder for the 

 flower while being measured. If the flower is picked off, gently 

 slipped into the notch and pulled down until the flattened corolla 

 lies upon the surface of the zinc plate, the ruler can be laid 

 firmly on the face of the corolla, at the desired angles, and the 

 two measurements taken very readily and after some experience, 

 sufficiently exactly. Other devices such as a squared sheet of 

 smooth paper ruled in millimeters, upon which the flower could 

 be pressed and the diameters read off, were found to be less 

 accurate than the more simple apparatus described above. 



On an average two or three flowers can be accurately 

 measured a minute. Considerable delay was caused by the 

 necessity of frequently washing off in alcohol the heavy gum 

 from the stems of the tobacco plants, which covered the hands 

 and instruments to such an extent as, at times, to render fur- 

 ther work impossible. An assistant recorded the measurements, 

 thus simplifying the work greatly. The records were so taken 

 that it is possible to refer, on a given date, to the measurements 

 of flowers on any particular plant in the whole series and, in a 

 general way, to recognize the region on the plant upon which 

 any single flower occurred. All measurements were made be- 

 tween 8 and 10 a.m. with the exception of November 9 and 13, 



