94 Error in Genera and Species. [zoe 



not invalidate Gray's genus Howellia, which may however be 

 the same species, its emersed fruit being still unknown, or it 

 may be one of the obscure aquatic Lobelias. 



The small annual Lobeliaceous and Campanulaceous plants 

 seem to possess a high degree of variability. The type of H. 

 limosa had expanded corollas, but in the numerous specimens 

 I have seen growing the flowers were cleistogamous. 



Campanula exigiia? which must be very near C. Reverchoni* of 

 which the mature fruit was not known, belongs to a group of 

 small annual Campanulas which have their home about the 

 Mediterranean. It is almost certainly an immigrant, may even 

 be C. Laflingii Brot., of which I have seen only an imperfect 

 specimen. The recently described C. angustifolia h is perhaps 

 only a semi-cleistogamous form, for in the thousands of plants 

 observed I never saw an expanded corolla The plate, where 

 this form is figured in comparison with C. exigua, is somewhat 

 misleading, the unexpanded style of the latter being figured in 

 relation with the fully evolute one of C. angustifolia. In the 

 genus the style before maturity is thick, cylindrical or clavate 

 and densely hairy; after expansion of the stigmatic lobes the style 

 appears to shrink to a third of its former thickness, and its 

 hairiness is much less perceptible. 



Another plant which may perhaps be a wanderer is Lithosper- 

 mum glabrum? Dr. Gray compares it with the Old World L 

 mcrassatum Guss., which I have not seen, but inspection of a 



hreviores, 2-2'/ 1 in lonri deZ,nnn f; f Peduncuhferecapillacei, primum flore 



Calvcis tubus angustus lKHn krt£*J? ^ eapsulam maturam xquantes, fere 4 lin longi. 

 erves in fructu ma ores Corolla 1 EkS -*' e tna "S"' ari lanceolati, V lin. longi. trin- 

 »on quinquepetauflobi d^ uwE '"• /T*"*' d ° rS ° , fi *' Sa ' 'l"i"quepartita, sed 

 auctions, inferiories laUorel ™£2«? ««£» ^ITVa **P«fto angnstiores et 



anther* connate paulinlum i, cnrv't "'■ £ f • rotul \ datl Filament* monadelpha; 



crassum, Wlobum Cam n ™,?, ' S" fer *S re8 dua *et«lo»0-ari«tat*. Stigma 



truncata ap « f ssa «X£fa^&f™J^ 8ma V C "' 3 5* ,in ,0n * a < ' H " «£ "£ 

 oblonga, !utTOft,£a7nUi!fc Tsa? m&gn^ ,ySpem »' !*■«»«■ *Mb«M, parietalibus. Semina 



Kapsel von den Capachen Mezlerien^b Z?-£ ^"S*!. ,n,t der S ' ,itze "ervorragende 

 undnachdem Prodromus von De Candon* Z ?t ^ dlC Mor > a ^'Phi S chen Filamente. 

 haben. Dennoca glaubeich kaunTda^dt ™™ k' 6 Me ^rien halbkugelige Kapseln 

 toche Trennung rechtfertigen diirfien I - R A * St? ebene » y^inedenhei.en eine gener- 

 ics Rattan, Bot. Gaz 11 339 ' A " **"*"*, Bot. JBeitUHg. xxii, 217. 



i 4 -l V- ra -y' * y, i FI - "'• i >i - >• s»pp- »*. 



5JBastwood, Proc. Cal. Acad. ser. 3. i 132 m x ; 

 (OJ Cray, Proc. Am. Acad. xvi. 227. ' ' 



