ON LYSIMACHIA CUSPIDATA. 235 



was asserted by the late Prof. Miquel to differ entirely from the 

 Javanese plant. I left it, however, under the name assigned it by 

 the Hamburg botanist, because I possessed no authentic example 

 of the species referred to by Miquel. I have since, through Dr. 

 Scheffer's courtesy, been favoiu'ed with a sx3ecimen, from the 

 Buitenzorg Herbarium, of Blume's plant, labelled, my corre- 

 spondent believes — though he is not quite certain of this — in the 

 author's handwriting. From an examination of this, it is clear 

 that, as Miquel states,* Blume's L. cuspidata is identical with the 

 plant erroneously described and figured by Klatt t as L. uliginosa, 

 Bl. This, with L. evalvis, Wall., L. alter nifolia, Wall, (between 

 which two it should apparently stand), L. adoensis, Hochst. & Steud., 

 and L. Linum-stellatum, Linn., form Klatt's section Asterolinon. I 

 confess there seems to me no adequate grounds for the retention 

 of Link and Hoffmansegg's genus ;| but those who consider the 

 diminutive petals and [usually only) much fewer seeds as sufficient 

 to keep the two last-named species apart, and who think Klatt's 

 divisions — as I do not — generally natural and well limited, might 

 perhaps reserve for the others § the sectional name of Pseiidastero- 

 linon. It certainly does appear impossible to remove Pelletiera 

 verna, A. St. Hil., judging from a comparison merely of St. Hilaire's 

 own figures, jl from L. Linum-stellatum, Linn., of which indeed 

 Baudo surmised it might be only a variety ]*!i though it is acknow- 

 ledged by Klatt, as well as by Endlicher and Duby. But when Sir 

 Joseph Hooker, recognising this, places them together under 

 Asterolinon, thus assigning to this very unstable group species 

 with a 5-lobed corolla and others with three distinct petals, it is 

 perplexing to trace the logical process by which he has been 

 induced to resuscitate Duby's Apochoris, which has really no smgle 

 character but free petals to distinguish it from Lysimachia, a 

 large and ubiquitous genus, embracing plants with considerable 

 diversity both of habit and floral structure, and with which, as it 

 seems to me at least, Klatt with good reason re-combined it."^'* 



* 'Ann. Mus. Lugd.-Bot.,' iv., 144. 



+ ' Der Gattung Lysimachia,' 39, t. 23. 



I Which, by some oversight, Le Maout and Decaisne station in the tribe 

 AnagallidecB, distinguished by its circumsciss capsule, or pyxis. (' Traite Eiem. 

 de Bot.,' 219.) 



§ I find the stamens of L. evalvis distinctly connate into a short free cup or 

 ring; those of L. alternifqlia I have not examined: in both the seeds seem, 

 judging only from dissection of the ovary, quite numerous. 



II ' Ann. Sc. Nat.,' 2e. ser. xi. t. 4, especially the abnormal petal detected by 

 Decaisne, which presumably arose from cohesion, the normal condition in 

 Lysimachia. 



^ ' Ann. Sc. Nat.,' 2e. ser. xx., 850. 



** There are one or two inaccuracies in the revision of this Order in the 

 * Genera,' which testify to the work having been performed with great haste, 

 doubtless under the pressure of multifarious engagements. E. gr. Primula is 

 stated (ii., 631) to be divided by Duby into two sections; whereas Duby admits 

 five, and Endlicher the two named. Euprecht had perhaps a better conception 

 of the natural subdivisions of Primula than any other writer; if, however, 

 Schott's two sections, approved by Hooker, be admitted, they are undoubtedly 

 insufficient; and a third, SphondyHa, agreeing with the first in the ptyxis of the 

 leaves, and with the second in the abbreviated fruit, must be intercalated. 



