322 CLEISTOGAMIC FLO WEES. Chap. VIU. 



remaining five contained on an average lO'O seeds 

 per capsule. This is rather above the average 9-2, 

 which eleven capsules from perfect flowers fertilised 

 with their own pollen yielded, and considerably above 

 the average 7"9, from the capsules of perfect flowers 

 fertilised with pollen from another plant; but this 

 latter result must, I think, have been accidental. 



Hildebrand, whilst searching various Herbaria, ob- 

 served that many other species of Oxalis besides 0. 

 acetosella produce cleistogamic flowers;* and I hear 

 from him that this is the case with the heterostyled 

 trimorphic 0. incarnata from the Cape of Good Hope. 



Oxalis {Biojphytum) sensitiva. — This plant is ranked 

 by many botanists as a distinct genus, but as a sub- 

 genus by Bentham and Hooker. Many of the early 

 flowers on a mid-styled plant in my hot-house did not 

 open properly, and were in an intermediate condition 

 between cleistogamic and perfect. "Their petals varied 

 from a mere rudiment to about half their proper size ; 

 nevertheless they produced capsules. I attributed 

 their state to unfavourable conditions, for later in the 

 season fully expanded flowers of the proper size ap- 

 peared. But Mr. Thwaites afterwards sent me from 

 Ceylon a number of long-styled, mid-styled, and short' 

 styled flower-stalks preserved in spirits ; and on the 

 same stalks with the perfect flowers, some of which were 

 fully expanded and others still in bud, there were 

 small bud-like bodies containing mature pollen, but 

 with their calyces closed. These cleistogamic flowers 

 do not differ much in structure from the perfect ones 

 of the corresponding form, with the exception that 

 their petals are reduced to extremely minute, barely 

 visible scales, which adhere firmly to the rounded 



• ' Monatsbericht der ikad. der Wiss. zu Berlin,' 1866, p. 369. 



