OYNCECIUM. 399 



while in a flower of Catalpa a solitary perfect stamen, and 

 a complete absence of the sterile ones usually present, 

 have been observed. This might have been anticipated 

 from the frequent deficiencies in the staminal whorl in 

 these plants under what are considered to be normal 

 conditions. Reduction of the staminal whorl is also 

 not imfrequent in Trifolium repens and T. hyhridum, and 

 has been seen in Delphinium, &c.^ 



Meiophylly of the gynceciimi. Numerical inequahty in 

 the case of the pistil, as compared with the other 

 whorls of the flower, is of such common occurrence, 

 under ordinary circumstances, that in some text-books 

 it is looked on as the normal condition, and a flower 

 which is isomerous in the outer whorls is by some 

 writers not considered numerically irregular if the 

 number of the carpels does not coincide with that of 

 the other organs. 



But in this place it is only necessary to allude to devia- 

 tions from the number of carpels that are ordinarily found 

 in the particular species under observation. As illus- 

 trations the following may be cited : Arenaria tetra- 

 queta, which has normally three styles, and a six-valved 

 capsule, has been seen with two styles, and a four or 

 five-valved capsule. Moquin relates an instance in 

 Pohjgala vulgaris where there was but a single carpel, 

 a condition analogous to that which occurs normally 

 in the allied genus Mozinna. Beseda luteola occasion- 

 ally occurs with two carpels only, while Aconites, Del- 

 phiniums, Nigellas, and Paeonies frequently experience 

 H like diminution in their pistil. 



In a flower of Paimver Bhceas the ^vTiter has recently 

 met with an ovary with four stigmas and four parietal 

 placentae only, and to Mr. Worthington Smith he is 

 indebted for sketches of crocus blooms with two, and 

 in one instance only a solitary carpel. 



Moquin cites the fruit of a wild bramble (Ruhus) 



' Cramer, ' Bildangsabweich,' p. 90. 



