538 



Rydberg: Notes on Rosaceae 



visable to give it a distinct name. In many respects it resembles 

 strongly 0. altemans, and the author has entertained the thought 

 that it might be a 2-carpellary form of that species. The styles in 

 O Hapemanii are, however, more or less spreading, while in 0. 

 alternant the style is erect, showing that the latter is related to 0. 

 malvaceus. Besides the type, the following specimens may be re- 



ferred to 0. Hapemanii: 



Wyoming: Hartville, July 20, 1894, Aven Nelson 498. 

 The following specimens are referred to 0. altemans: 



Wahsatch Mountains, J 



1 159 



J 



Mountains, S. Watson 



M 



A. A. Heller adopted the name 0. paueif 



malvaceus. 

 flora Nutt. 



Spiraea pauci 



paucift 



ipidi 



folia paucfl 



This variety was, 



however, primarily based on S. monogyna Torn, and hence e ong 



rather to 0. monogynus. 



Spiraea 



Spiraea parvifolia Benth. is a very strange and interestmg J^ 

 cies. Maximowicz referred it to the section (now the gei^ 

 Petrophytum. It is, however, a true Spiraea, of **~*^ it 

 the racemose inflorescence and the entire leaves. e ^ 



the depressed habit nor the fruit of Petrophytum. As n ^ 

 known to me, it has been collected but once. By th< \ CC ? r Botan _ 

 the director of the Kew Gardens, England, the New Yor ^ 

 ical Garden has received an excellent drawing of the J^^ 

 some fragments of the plant, enough to show its rea ^ 



Unfortunately, 5. parvifolia Benth. is antedated h J ^^ 

 Raf. I therefore took pleasure in naming the plant ^. 



after the discoverer. 



Flora 



The species which appeared in the North m ^ f 



under the name 5. Steveni has had a peculiar history. ^ the 

 botanists referred it to 5. chamaedryfolia L., a species ^^ 

 flowers in simple corymbs, instead of in flat-topped pai g en( lcn$> 

 it was referred, together with S. corymbosa, S. luctda, 



