45cS 



The Journal of Heredity 



From the size of anthers and sti<i;ina, 

 colors and habit of growth it would 

 seem that the hybrida had been univer- 

 sally used for both male and female 

 parents, until recent years. 



Though advised otherwise, in m\- 

 earl\' work I chose the form known as 

 grandiflora as the female parent for my 

 doubles, ])robably because the flowers 

 were easier to work with. Later when 

 an ideal had formed itself in my mind 

 the grandiflora seemed more likely to 

 give the desired results. 



Many doubles have rudimentary or- 

 gans of re])roduction, but in my first 

 work in 1901 I noticed this and formed 

 the habit of examining each bloom care- 

 fully before ]>icking it to ]jollenize with. 



The first perfect double bloom was 

 found on a hybrida plant in 1910. This 

 plant had dcHcately fluted flowers with 

 cream colored pollen and, when pollen- 

 izcd with another flower from the same 

 ])lant, matured a capsule of seed. The 

 stamens of this flower were many, 

 rising directly through the center, the 

 filaments l)eing Ixjund together by a 

 band or collar, while the ovary sat 

 upon a torus. The ovary of the single 

 form sits directly upon the calyx. 



From this capsule of seed thirt\'- 

 seven j^lants were raised. No records 

 were kept until 1911, but as near as I 

 can remember about 75% were double, 

 h)oth single and double being of the 

 hybrida type. None of these plants 

 gave many perfect flowers though all 

 were examined for reproductive organs 

 and some .seven or eight matured seed. 



That .season among our regular 

 doubles a])i)eared a semi-double of 

 steel blue and white which bore all 

 ])erfect flowers, and on an inferior 

 d(juble red was found a capsule of 

 seed which had matured without hand 

 pollenizing. 



From these three distinct types 660 

 plants resulted in 1912, 85% doul)k' 

 and 22%, seeding slightly. 



CKH.VT SARI.AHILITV 



The ])etunia is jx-rhaj^s the mo.st 

 variable flower under cultix'ation but 

 its fluctuations have a certain regularity. 

 In this generation, the three types 

 being planted together, the wildest con- 



fusion prt\aik(l. In it apjicartd for 

 the first time the true grandiflora, 

 represented b\- three plants of deep 

 magenta color with steel-blue ])ollen. 

 Their doubling was of an entirely 

 different nature, — all extra petals were 

 adherent to the corolla tube instead of 

 the usual mass of j^etals and stamens 

 which generalh' fill the center of the 

 flowers. Nearly all blooms on these 

 plants were perfect, thovigh they did 

 not all mature .seed. 



In this planting were also some very 

 small inferior doubles of dingy purple 

 flowers, which were perfect with the 

 same manner of doubling and which 

 matured several capsules of .seed without 

 pollenizing. Both extremes ha\'e the 

 same form and both are fertile. 



From the grandiflora crossed by the 

 hybrida double and some seed of the 

 hybridas also we raised in 1912, 510 

 plants, 73% in double, 25% seeding. 

 Many of the flowers showed great 

 variety of color, beauty and size. 



In 1913 we got 187 plants with 73% 

 double but 33% seeding. This year 

 marked a decided change, the grandi- 

 flora heretofore recessive became the 

 dominant type with blooms of extraor- 

 dinary size, while its seeding capacity 

 had increased 8%. Both beauty and 

 rcproductiveness had developed to such 

 an extent that for our stock seed I 

 crossed a perfect double with a perfect 

 double for the first time, but disaster 

 overtook me for our seed beds with our 

 entire stock of seedlings were washed 

 out by the floods of 1914. 



Replanting from our selling stock 

 yielded 918 plants, 85% double, 42%o 

 seeding. 1915 jiroduccd 567 plants, 

 90^;^ double, 40^;;, seeding. While the 

 increase of doubles has been quite 

 steady the seeding precentage has not 

 increased so rapidly due to the vise of a 

 ])lant that carried .singleness in its 

 jjollen but with other qualities which I 

 wished to preserve. 



I have not yet made the recijirocal 

 cross again but exijcct to do so this 

 season, as many of the fine large flowers 

 are ])erfect, seeding as freely as singles 

 when ]K)llenized. 



Doubtless my work would ha\e been 

 done on entirely different lines had my 



