169 



i^M-ains of carbohydrate matcrials. The cystocarps described for the 

 gcnus by J. C>. Agardh (Ioc. cit.) are thosc of Neurocaulon reniforme, 

 so that the First description of the cystocarps of Constantinca proper 

 is that of ScHMiTz and Hauptfleisch which is excellent. Their mate- 

 rials may have been of C. rosa marina, but possibly may bave been 

 the plants of Farlow, Anderson, and Eaton, sent out as C. Sitchensis, 

 but now to be separated as C. simplex. Freeman (Ioc. cit.) failed to 

 lìnd cystocarps in C. subulifera and the writer has been equally 

 Linsuccessful even in the abundant material at bis disposai. The wri- 

 ter, howcver, has found procarpic structures in this last species. Cy- 

 stocarpic material of C. simplex is abundant and have been briefly 

 described by the writer (Ioc. cit.). 



The cystocarps of C. simplex, the only species in which the 

 writer has been able to study these structures, correspond exactly 

 to those described by Schmitz and I1auptf[,eisch. They are reniform 

 masses of spores, situated in the cortcx of the upper side of the la- 

 mina, each in a cavity without special enveloping tissue, and ope- 

 ning outward (and upward) by a short pore. Projecting up into the 

 center of the spore mass is a row of cells forming a stalk or pedi- 

 cel, from the tip of which, the gonimolobes spread in ali directions. 

 The gonimolobes develop equally and simultaneously and, while so- 

 mewhat distinct at first, are soon confluent. In the cross section of 

 a cystocarpic lamina, there are to bc seen the innumerable short, 

 more or less curved and moniliform auxiliary branchlets so characte- 

 ristic of members of the Dumontiaceae. Tlie cystocarps occupy a 

 dense zone, centripetal in development, beginning to form at the 

 periphery and proceeding towards the center. A few plants of C. subu- 

 lifera show the characteristic auxiliary branchlets which, however, are 

 not uniformly distributcd as in C. simplex, but occur in small groups. 

 No antheridia have been detected in any of the species. 



From the above considerations, the foUowing conclusions may 

 be dravvn : 



i). That the genus Constantinca is to be restricted at present, 

 to the three species, C, rosa maritia, C. simplex, and C. subulifera : 

 2). That Constantinca, thus constituted, is a genus of the Du- 

 montiaceae in the sense of Schmitz and Hauptfleisch, characterized 



* 



