99] NORTH AMERICAN AND WEST INDIAN CUSCUTA—YUNCKER 9 



Shortly after the appearance of Choisy's paper Engelmann, in 1842, 

 published his Monography of the North American Cuscutineae. He 

 treated the group as a tribe of the Convolvulaceae, dividing it into two 

 genera: (1) Cuscuta, with a gamosepalous 4-5 parted calyx and (2) Lepi- 

 danche, with 10-15 imbricated sepals. He described seven species of 

 Cuscuta and one of Lepidanche, all of which he considered as new, but in 

 a later paper he treated five of them as synonymous with previously 

 described species. 



In 1845 the ninth volume of de Candolle's Prodromus, in which the 

 Convolvulaceae were treated by Choisy, made its appearance. Choisy 

 here considered the dodders as a tribe of the Convolvulaceae and included 

 forty-nine species. He included Engelmann's species, but considered 

 most of them as doubtful or relegated them to synonymy. 



Pfeiffer in the meantime had become interested in the dodders coming 

 under his attention and, later in 1845, published a revision of the group. 

 He did not believe it to be subordinate to but rather coordinate as a 

 family with the Convolvulaceae and so treated it. He divided the family 

 into three genera on the basis of stigmatic characters, as follows: (1) 

 Cuscuta, with linear stigmas, (2) Epilinella, with clavate stigmas, and (3) 

 Engelmannia with capitate stigmas. The following year (1846) he 

 elaborated his original discussion and included Engelmann's genus Lepi- 

 danche. 



Buchinger (1846) in reviewing Pfeiffer's article showed that his Engel- 

 mannia is untenable because of previous occupancy and suggested the 

 substitution of Pfeifferia. This too, however, had been previously used 

 to designate a genus of the Cactaceae. As early as 1790 Loureiro in his 

 Flora Cochinchinensis had established a monotypic genus which he called 

 Grammica. This genus was apparently unknown to Pfeiffer and Buch- 

 inger. Later the species included in this genus was shown to be a member 

 of the group of Cuscutas possessing capitate stigmas. It would appear 

 proper, therefore, to use the name Grammica to designate the members of 

 any sub-group possessing similar characters, and Engelmann in a later 

 paper uses it in this manner to designate those species with capitate 

 stigmas. 



Des Moulins published his Etudes organiques sur les cuscutes in 1853. 

 He, like Pfeiffer, considered the group as an independent family and 

 divided it into two tribes using the character of the circumscission of the 

 capsule as a basis for the division. He showed that it is possible to further 

 divide these tribes into five genera on the basis of the characters of the 

 seeds, the capsules or the styles. He decided to use the styles as showing 

 the best differential characters and proposed the following arrangement: 



Capsule circuniscissile. Tribe Cuscuteae 

 1 . Styles filiform; seeds not winged Cuscuta 



