ON PrjMARY CHARACTERS IN CERASTIUM. 9 



The most characteristically defined group in Cerastium includes 

 those species in which the teeth of the capsule after dehiscence are 

 finally circinate-convolute. In none of the allied genera does this 

 peculiarity in the capsular teeth occur. But it is to be noted that 

 in all the species referable to this group the ripening capsule, while 

 increasing in length, does not become curved, except in C. Arme- 

 niacum, where it is considerably curved, and more than half exserted 

 from the calyx. If, without regard to the number of teeth (whether 

 six or ten), those species in which the teeth are finally circinate- 

 convolute are included in the group defined by Seringe as the 

 section Strephodon, there only remain in the group defined by 

 Bartling as the section Dichodon two well-known species, C. tri- 

 gynuin and C. anomalum. To these may, however, be added C. 

 m.auritanicum Pomel, and C. melanandrum Maxim. 



Examination of the capsules of other species shows also that 

 the number of styles in specimens is an inconstant character. A 

 well-defined American species, C. Texanum, has been described by 

 Dr. N. L. Britton in which the styles vary 3-5, but the capsular 

 teeth are invariably circinate-convolute. The same obtains in 

 C. indiciuii W. & Arn. In Wight's specimens I have noticed several 

 ovaries with only three styles instead of five. Mr. Henry Trimeu, 

 in his Flora of Ceylon, says of the specimens which he examined, 

 " Styles 5, not 3,'' as given in the Fl. of Brituh Lidia." 

 This is certainly not stated in FL of British India, i. p. 227. 

 Again, in a Siberian species, C. obtusifolium Kar. & Kir., included 

 in the section Strephodon,\ I find on the examination of authentic 

 specimens that there are only three styles, and that the six capsular 

 teeth are subrevolute at the margin, and not circinate-convolute at 

 the tip, but straight and obtuse : so that but for the form of the 

 petals (should the character be sufficiently distinctive) the species 

 might as well be reduced to a variety of C. trigijnum. 



It may further be pointed out that in C, trigynum and allied 

 species the capsule is never curved, which serves with other charac- 

 ters taken into consideration to separate them from the large group 

 of which G. glovieratum and C. arvense are well-known examples ; a 

 group in which the character of the 10-dentate capsule is normal 

 and less inconstant. 



In the matter of habit and aspect Stellaria includes slender, 

 diffuse, glabrous herbs, while the species of Cerastium are pubescent, 

 though C. perfoliatum L. is a conspicuous exception, and the per- 

 ennial forms are often tufted. On the other hand, two species of 

 jfirenaria with long cylindrical capsules, viz. A. Guicciardii Heldr. 

 and A. purpnrascens Eamond, may be regarded as connecting links 

 with the normally 3-styled species of Cerastium. In Cerastium the 

 capsule may be said to be always more or less elongated, while in 

 Stellaria and Arenaria it is characteristically ovoid or oblong. 



An important character noted by Fenzl| is that the seeds of 

 some species are angular (in the dried state), from shrinking of the 



* There is probably also au error in transposition of figures here, 

 t Ledeb. Fl. Rossica, i. 398. J Ledeb. Fl. Rossica, i. 415. 



