FACTORS INFLUENCING FLOWER SIZE IN NICOTIANA 347 
Under the general heading "age of plant factor" the question of 
the differences in size of flowers on plants growing the second year on 
their own roots and flowers on the same plants the first year is included. 
On this point the data is given in Tables IVa, IVb, Va, and Vb, and 
in condensed form in Table VII. Two facts are strikingly evidenced 
Table VII 
Influence of Different Years on Corolla Measurements on the Same Plants 
Designation 
Me 
1912 
an 
1913 
Difference 
Fi Has : spread 
35-8i±.i5 
39.57±-32 
3-76d=.35 
52.95±-i6 
54.57=t.i6 
I.62d=.24 
Fi H44: spread 
33.11 ±44 
3942=t.52 
6.3i±.68 
length 
52.16dr.46 
5342±.37 
i.26±.6i 
by the results as given. First, the spread of corolla of flowers which 
are produced on new vigorous laterals from plants which have been 
given an involuntary period of rest by cutting back is distinctly larger 
than the spread of flowers measured toward the end of the first season 
of growth and flowering. These differences are distinct and significant 
when compared with their probable errors. The second significant 
point illustrated above has to do with the relatively slight difference 
between the length of corolla in the case of the two periods of measure- 
ment. A difference of 6.31 i .68 mm. in the spread of Fi H44 is 
associated with a difference of only 1.26 d= .61 mm. in the length of 
the same flowers and similarly for Fi H38 we have a difference of 
3.76 =b .35 mm. for spread as contrasted with 1.62 d= .24 for length 
of corolla. 
It must be borne in mind in this connection that the plants the 
first year (12 Fi H38 and 12 Fi H44) were measured during the last two 
weeks in October or toward the end of the period of measurement as 
given for the other groups of plants in 191 3, while in 1913 (13 Fi Hss 
and 13 Fi H44) the flowers were measured early in their second period 
of blooming. In other words, if in 1913 the flowers had been measured 
during October, their size would undoubtedly have been more nearly 
equal to that given by the measurements in 191 2. The dates of 
measurement and results as they stand are, however, particularly 
significant for our purpose. There is obviously, also, a decided dis- 
crepancy in the number of measurements taken in 191 2 as compared 
