134 University of California Publications in Botany [Vol. 5 



In reference to the diameter of the largest flower measured 

 in variety II there is doubt about extending the limit to 23 mm., 

 since but one record of this diameter was noted and the figure 

 "3" is somewhat obscure in the original record. It may be 

 claimed that a degree of variation in corolla diameter as slight 

 as that noted in variety III — 13 mm. to 15 mm., or 3 mm. — or 

 even in variety II of 5 mm., is impossible. A superficial glance 

 at the plants in flower would certainly seem to substantiate such 

 a claim. The diameters given, however, have reference to 

 measurements made upon flowers of the same age or, better, in 

 the same stage of development. Only such flowers were measured 

 as exhibited all the anthers shedding fresh pollen. It is not 

 a difficult matter to decide by a far from critical examination 

 whether or not this condition is present in a given flower and, 

 after a considerable acquaintance with the flowers of N. acumin- 

 ata, it has been found possible to decide very quickly. 



All the measurements in 1910 were made between 7 :45 and 

 9 :30 a.m. The newly opened flowers often show the corolla lobes 

 fully extended and flattened but the anthers still green. There 

 is a certain characteristically immature look about such a flower 

 very easy to recognize at a glance. In such a flower on a bright, 

 warm day the anther lowest in the tube will usually break during 

 the afternoon and on the following morning all the remaining 

 four will be shedding pollen. I have found that the difference in 

 corolla diameter in these two conditions usually exceeds 2 mm. in 

 all varieties of N. acuminata. Thus adding 2 mm. to the corolla 

 diameters for the two extremes given for variety I we might 

 have on the same plant flowers measuring 24 mm. and 30 ram. 

 and all sizes intermediate. Likewise for variety II, 17 and 

 24 mm. and for variety III, diameters measuring 11 and 17 mm. 

 and all diameters between. If Ave add, again, the appearance 

 of partially opened buds and flowers past anthesis, there 

 would seem to be on a plant, for example of variety II, the 

 greatest individual fluctuation in corolla diameter. Still it may 

 be seen that, allowing 2 mm. l^eyond either extreme of fluctua- 

 tion in corolla diameter of fully opened flowers as noted in the 

 above table, there is yet a definite division into the three 

 varieties — i.e., a large (24 to 30 mm.), a small (11 to 17 mm.). 



