R"iT)BERG: Notes on Rosaceae 143 



related to R. jamaicensis, but without the white tomentum 

 on the lower surface of the leaves and being in all respects 

 less pubescent. This has the same habit as R. jamai- 

 censis, being reclining over bushes. Grisebach compares it 

 with R. fagifoliiis. This was probably the reason why Focke 

 transferred it to his subgenus Lampohatus and associated it with 

 R. fagifolius and R. scandens. For my part it seems as if the 

 Jamaican plant was unknown to Dr. Focke. Under any circum- 

 stances, the plant which I take for R. alpinus is the same as R. 

 guyanensis Focke. At least I have not been able to find any 

 differences. A photograph and some fragments of the type of R. 

 guyanensis were kindly sent me from the Botanical Garden at 

 Berlin and these are exactly like Shreve's specimens of R. 

 alpinus from Cinchona, Jamaica. Other specimens from Guiana 

 are the same. Otto Kuntze's specimens from Venezuela are similar 

 but more pubescent, at least on the veins of the under side of the 

 leaves. Specimens collected near Cinchona, Harris Q141, have 

 the leaves more pubescent beneath, and may be hybrids with R. 

 jamaicensis. 



Rubus vulcanicola (J. D. Smith) Rydb. is closely related to 

 the preceding and mainly distinguished by the mostly five-foliolate 

 and more shining leaves, and longer pink or rose petals. I have 

 seen the following specimens: 



Costa Rica: Volcan Poas, March 1896, John Donnell Smith 

 64QS; 1896, Tonduz lOQoy; Rancho del Achiote, Volcan Poas, 

 1889, Pittier 806; Volcan Barba, 1890, Tonduz 2120. 



Sapidi 



Rubus Nelsonii Rydb. is probably related to R. sapidus but 

 differs from that species as well as from its other relatives in the 

 narrower, long-acuminate leaflets, and also in its terete stem. It 

 is known only from the two sheets of the type collection. 



Rubus amplior Rydb. is related to R. sapidus, but differs in the 

 terete stem, the dark green and firm leaflets, these being less cordate, 

 and having broader teeth. Known only from the type collection. 



Rubus Palmeri Rydb. is closely related to R. sapidus, but the 

 leaves are lighter green, the inflorescence somewhat glandular, as 

 well as villous-tomentose and the teeth of the leaves are broader 



