348 PROCEEDINGS OP THE AMERICAN ACADEMY 



Gaura and its near allies. While so engaged, I found that Linnaeus 

 had noticed these appendages at least as early as in the fifth edition of 

 his Genera ^ Plantarum.* It is probable that they were mentioned 

 in his paper in the Acta Holmiensia, in which he established the 

 genus. With the exception of Jussieu,t subsequent authors seem to 

 have lost sight of them. 



These appendages are present in all the species of Gaura (though 

 in G. jyarvijlora reduced to a minimum) ; but are wanting in Stenosi- 

 phon and the G. heterandra of Torrey. On account of this and other 

 differences, I have been led to propose the latter as the type of a 

 new genus {Heterogaura). 



The nearest analogue to these appendages with which I am ac- 

 quainted will be found in many of the ZygophyllacecB, — Larrea, for 

 example, — where they are much more developed, and are almost as 

 long as the filament. From their position, it may be inferred that 

 they do not represent an additional set of stamens, but are part of 

 the filament to which they are attached. If the hypothesis of chorisis 

 be admitted, they may be regarded as fine examples of it, either of 

 vertical chorisis, if their position before the filament be taken as nor- 

 mal ; or of collateral, if (as the notch at their apex sometimes may 

 indicate) they be considered as homologous with a pair of stipules. 

 I have never been able to find such appendages on the inside of 

 any broad-based filaments. 



Though the presence or absence of the appendages is a good generic 

 character in Gaurinece, they are much too variable to serve for limit- 

 ing species, except in G. parvijiora, where they are reduced to mere 

 papilla3, and in G. Drummondii, which has those of the filaments oppo- 

 site the petals decidedly shorter than the others. 



As to the indusium or ring surrounding the base of the stigma, in 

 the same genus, I found that this also had been noticed before, or at 

 least represented. That is, it had been figured in Payer's Organo- 

 genie Vegetale (tab. 150, f. 34-36), in young flower-buds of G. bien- 

 nis ; but there is no reference to the subject in the letter-press. 



It is clearly of the same nature as the indusium in Ericacece, noticed 

 by Robert Brown, in which, as Mr. Bennett remarks, the likeness to 



* " Glandula nectarifera conica intra basin singuli." — Linn. Gen. Plant. Ed. 5 

 (1754). 



t " Glandulas 8 in G. hienni memorat Linnyeus summo calici adnatas." — Juss. 

 Gtn. Plant, 



