Chap. V. HETEEOSTYLED TEIMOEPHIO PLANIB. 193 



matter, without any good pollen-grains, and they never 

 shed their contents ; they were in the state designated 

 by Gartner * as contabescent, which term I will for the 

 future use. In one flower all the anthers were conta- 

 bescent excepting two which appeared to the naked 

 eye sound ; but under the microscope about two-thirds 

 of thepollen-grains were seen to be small and shrivelled. 

 In another plant, in which all the anthers appeared 

 sound, many of the pollen-grains were shrivelled and 

 of unequal sizes. I counted the seeds produced by 

 seven plants (1 to 7) in the first lot of eight plants, 

 probably the product of parents fertilised by their 

 own-form shortest stamens, and the seeds produced by 

 three plants in the other two lots, almost certainly the 

 product of parents fertilised by their own-form mid- 

 length stamens. 



I'lant 1. This long-styled plant was allowed during 1863 to 

 be freely and legitimately fertilised by an adjoining illegitimate 

 mid-styled plant, but it did not yield a single seed-capsule. It 

 was then removed and planted in a remote place close to a 

 brother long-styled plant No. 2, so that it must have been freely 

 though illegitimately fertilised; under these circumstances it 

 did not yield during 1864 and 1865 a single capsule. I should 

 here state that a legitimate or ordinary long-styled plant, when 

 growing isolated, and freely though illegitimately fertilised by 

 insects with its own pollen, yielded an immense number of 

 capsules, which contained on an average 21 • 5 seeds. 



Plant 2 This long-styled plant, after flowering during 18G3 

 close to an illegitimate mid-styled plant, produced less than 

 twenty capsules, which contained on an average between four 

 and fivo seeds. When subsequently growing in company with 

 No. 1, by which it will have been illegitimately fertilised, it 

 yielded in 18C6 not a single capsule, but in 1865 it yielded 

 twenty-two capsules : the best of these, fifteen in number, were 

 examined ; eight contained no seed, and the remaining seven 

 contained on an average only three seeds, and these seeds were 



• ' Bcitrage zur Kenntniss der Befruohtung,' 1844, p. IIU. 



