834 BULIKTIN DK 1,'hKRBIKU BOISSIKR (2'"" SÉR.). 1907 (5) 



Maximowicz (in Mel. Biol. FX. 50) says lie lias coliected spécimens of this 

 variety of Ihe species « mox humilia granditlora calycis laciniis acuminalis, 

 inox elala floribus fere Iriplo minoribus calycis laciniis fere angusle 

 ovalis aculis, sed inter lalia erant simillinia floribns normalibus paullo 

 lanliira minoribus onusta. » This variable characler in examples from Ihe 

 same dislrictis one ol' Ihe reasons why I hâve hère uniled Fenzl's var. a 

 wilh var. ß lusus 1, — as Fenzl himself says (/. c. 395), « varielas a in 

 var. ß medianle lusu 1. heec in var. y mediante lusu 2, insensililer 

 transeunl. » Wilh Lhe exceptions mentioned, ail Ihe spécimens of var. 

 fasciculala hère cited are in Herb. Kew. 



The earliest référence lo lhe species is in Pallas's Accounl of his 

 Travels; and the following exlracl (which is daled in lhe margin 

 25 May 1772) is Iranscribed from lhe French translation of Gauthier 

 de la Peyronie (1793) : — « Parvenu à la moitié du chemin, je m'arrêtai 

 « dans une plaine sablonneuse où l'Onon forme une nouvelle sinuosité. 

 « Je fis halle pour faire manger les chevaux près du petit ruisseau de 

 « Schilboungou. Ici l'Onon est bordé de rochers escarpés, et surtout vers 

 « sa rive gauche. Je vis dans la plaine sablonneuse le cherler en forme 

 » de sedum, petite plante très commune en Daourie. » This cerlainly 

 refers to var. fasciculata, which is distinguished from var. uniflora in lhe 

 Prodromus by « floribus subumbellalis, caule elongato, foliis non 

 curvatis. » 



Iconogr. — Lebeb. le. pi. fl. Rossicœ. 401. This excellent plaie of the 

 lypical form is lhe only figure of lhe species known to me. 



The characlers relied upon by Edgeworlh for separating St. decumbens 

 from Ihis species are incorrect. In Fl. Brit. Ind. I. 234, he says, il is "a 

 very common and variable Alpine Himalayan plant, allied to S. petrœa 

 Bunge, but at once distinguished by ils shorter less rigid leaves. and 

 denser pubescence. " Référence to aiilhentic spécimens of both.and com- 

 parison of descriptions, show no such discrepancy. Edgeworlh gives the 

 lenglh of the leaves as V»-V^ inch, which is longer, not shorter, than 

 that given of the leaves of S. petraea (in lhe original description); nor 

 indeed is lhe pubescence more marked in the Indian plant, but an exa- 

 minalion of the available malerial shows Ihal lhe reverse is the more 

 frequently the case. These two characlers, trivial in themselves, are raosl 

 inconstant. Again, in var. 3, Edgeworlh wriles "sepals 4" but in lhe 

 Sikkim spécimens I examined only a few flowers seemed to hâve four 

 sepals. — and in thèse lhe fifth might hâve become delached. Again, 

 Siberian spécimens of "5. petrœa''' do nol show the leaves to be less 

 nilescenl than those of hidian examples of St. decumbens. Moreover, in 

 the description of the lalter, the accidenlal omission of lhe word leaves 

 somewhat obscures lhe sensé and meaning. 



2. St. arenaria Maxim. Fl. Tangutica, p. 91, t. 29, fig. 18 (1889). 



Replans ex nodis cauliculos 75 mm. apice in ramos pube crispa 

 bifariam puberulos solutos emitlens. Gaules primarii arena sepulli hori- 

 zontales teretiusculi fîlum emporeticum crassi glabri inlernodiis 25-36 mm. 

 folia ovata squamiformia ferentes, ex axillis fîbras radicales et cauliculos 



