OP AKTS AND SCIENCES : AUGUST 10, 1864. 349 



the more strongly developed one of Goodenoviece is too obvious to 

 escape attention.* In several species of Gaura, the whole structure 

 is strikingly like that of Eingcea repens, i. e. with the four (in Epigcea 

 five) lobes of a manifest indusium adnate to the under side of each 

 projecting stigma. Sometimes, as in G. villosa and .G. tripetala, the 

 lobes of the indusium are confluent with or firmly adherent throughout 

 to their stigmas. In others, as G. Lindheimeri, the lobes are free at 

 the summit, or soon become so, or may easily be separated down to 

 their base. In G. coccinea, the annular or crateriform indusium is 

 quite entire and mainly free ; in G. longijlora, nearly so ; in G. sinu- 

 ata, slightly four-lobed. In G. parvijiora alone, the stigmas are 

 scarcely exserted beyond the indusium. 



Probably the indusium may in some instances furnish good charac- 

 ters ; but they are difficult to fix in dried specimens, and vary some- 

 what during and after anthesis. G. Lindheimeri is the only species 

 of which I can at this season examine the live plant. 



In Stenosiphon, the indusium is peltate and nearly entire, but (in 

 the dried specimens) more or less plicate or intruded between the 

 stigmas, so as partly to enwrap their bases. In Gaura heterandra, 

 Torr., there appears to be no indusium : at least, if there is anything 

 answering to such an organ, it is so completely confluent with the 

 entire and disk-shaped proper stigma that it is not at all discernible. 

 This, added to the total want of appendages on the base of the fila- 

 ment, as already remarked, strengthens the conviction that this anoma- 

 lous plant is a new generic type, much more distinct from Gaura, 

 indeed, than Stenosiphon is. The habit of the plant, which is rather 

 that of Clarkia, and the equally spreading instead of the dechned 

 stamens, confirm this view. 



The three genera of the group {Sckizocarya, Spach, and apparent- 

 ly his Gauridium, not differing from Gaura) may be contrasted as 

 follows. 



Subtribus GAURINE^, Spach. 



Calyx tubo supra ovarium producto, post anthesin deciduo. Stamina 

 duplo petalorum numero. Ovarium (4-3-, rarissime 6-) ovulatum. 

 Fructus nucamentaceus, abortu 1-4-spermus. 



1. GAURA, L. Calycis tubus obconicus vel elongato-clavatus. 



* Plantse Javanicae Rariores, 1 . p. 86. 

 VOL. VI. 37 



