ALABASTRA DIVEBSA 405 



BORRAGINE^E. 



Ma 



i. p. 47, t. 13 ? 



Corumba ; Robert, 804. 



This genus was founded by Humboldt and Bonpland in 1808 

 (/. c.) upon a specimen from Loxa in Peru. The description, how- 

 ever, was faulty in many respects, and it was not till several years 

 after that Kunth's careful study of the fruit (Nov. Gen. et Sp. PL 

 vii. 207) gave a clue to the affinity of the plant. Much doubt still 

 remained respecting the genus, till in 1879 appeared Grisebach's 

 Symbola ad Floram Argent inam, and there (p. 270) he describes a 

 Saccellium, identified with S. lanceolatum Humb. & Bonpl., as 

 occurring in the province of Oran. After Kunth's careful ex- 

 amination, the chief uncertainty a propos of Saccellium concerned 

 the flower, Humboldt and Bonpland having described this as 

 polypetalous, with antipetalous stamens. Grisebach, however, 

 found the corolla to be gamopetalous, with the stamens inserted 

 in the throat ; and this statement I can confirm, although the 

 flowers at my disposal (they are very small at time of pollination) 

 have not been quite what one would desire. 



I am not sure that Grisebach is correct in referring the 

 Argentine plant to S. lanceolatum. To take the leaves alone, 

 which is almost all we have to depend upon, these organs are 

 lanceolate and six inches long in the type, whereas Grisebach's 

 plant, like M. Robert's under notice, has oblong lanceolate leaves. 

 The Gottingen professor gives no measurements ; the Matto Grosso 

 plant, in all probability conspecific with the Argentine, lias much 

 smaller leaves, only lf-2| inches in length. The name given 

 above must therefore be regarded as provisional. 



I have failed to find any recent record relating to this rare and 

 interesting genus, and, so far as I am aware, this is the first 

 announcement of a Brazilian habitat for it. But seeing that the 

 neighbourhood of Corumbi has been visited lately by several 

 zealous botanists, specimens of Saccellium most probably form 

 part of their collections. 



BlGNONIACEJE. 



Jacaranda rufa Manso, Enum. subst. braz. p. 40. 

 Sant* Anna da Chapada ; Robert, 419 and 709. 

 Flowers were gathered in July, fruits in November. 



Jacaranda (§ Monolobos) Roberti, sp. nov. Stirps humilis 

 e rhizomate sat crasso caules breves ascendentes piloso-pubescentes 

 et folia perpauca sese longe excedentia gignens, foliis bipinnatis 

 4-6-jugis pinnis oppositis vel suboppositis vel hac atque iliac 

 manifesto alternis imparipinnatis 5-20-jugis, foliolis oblongis vel 

 lineari-oblanceolatis apice mucronatis basi decurrentibus margine 

 revolutis coriaceis fac. sup. glabris inf. piloso-pubescentibus juveni- 

 libus aliquantulum bullulatis, panicula caulem terminante abbre- 

 viata angusta pauciflora piloso-pubescente, bracteis anguste lineari- 

 oblanceolatis pedicellis subaequilongis, calyce campanulato extus 

 piloso-puberulo intus glabro ultra medium in lobos late oblongos 



