1925) REHDER, NEW SPECIES, VARIETIES AND COMBINATIONS 203 



horizontalis Moench, though in habit it looks quite different; instead of 

 sending out long trailing shoots, it develops on rather short and often 

 twisted prostrate stems dense masses of crowded very short branchlets 

 forming irregularly subglobose clusters flattened at the apex and not more 

 than 15 cm. tall. The leaves are all scale-like, minute and closely im- 

 bricated, bright green except near the end of the branchlets where they 

 assume in autumn the purple-violet color peculiar to J. horizontalis. 

 Occasionally, but very rarely 



horizontalis 



this form. 



Juniperus horizontalis f. alpina, comb. nov. 



Juniperus Sabina [var.] 5. alpina Loudon, Arb. Brit. iv. 2499, fig. 2362 (1838). 



Juniperus alpina Loddiges Cat. (1836), as synon., ex Loudon, I.e. 

 A typo recedit ramis ascendentibus vel erecto-patentibus nee prostratis, 

 ramulis suberectis, caulibus initio suberectis et ad 75 cm. altis, demum 

 prostratis apice ascendentibus, ramos suberectos emittentibus, foliis fere 

 omnibus acicularibus oppositis, in axibus primariis robustioribus tantum 

 ternatis, erecto-patentibus lanceolatis vel lineari-lanceolatis 2-5 mm. longis. 



Specimens and plants examined: plants growing in the Arnold Ar- 

 boretum under no. 8028 (specimens: October 27, 1924). 



This form differs considerably from the type in its upright or ascending 

 habit and in its juvenile acicular foliage. Young plants have upright or 

 nearly upright stems with upright or ascending branches forming little 

 columns which finally bend over and become prostrate, but are always as- 

 cending at the apex and develop branches which also are upright or as- 

 cending. In one of the parks at Rochester there are plants more than 



Mr 



lengtl 



from 20 to 24 inches or occasionally to 30 inches in height. The plants 

 are not known to have produced fruit. I can not refer this form to any 

 other species than to J. horizontalis. The final habit of the plant is similar 

 to that of J. Sabina L., but the acicular leaves are only slightly spreading 

 or even loosely appressed, spreading at an angle of less than 45 degrees 

 and are more abruptly and less sharply pointed, while the juvenile leaves 

 of J. Sabina and also of J. virginiana are more gradually narrowed, stiffer 

 and more pungent and spread at an angle of 45 degrees or more ; also the 

 green midrib on the upper surface of the leaves is more pronounced in 



horizontalis. Further- 



odor 



J. Sabina exhale and the pale bluish violet bloom the plant assumes in 

 autumn would indicate that it belongs to J. horizontalis rather than to 

 J. Sabina, though the original name suggests a plant introduced from the 

 European Alps. The description of Loudon, however, as far as it goes, 

 and the figure of a branchlet agree well with our plant, which came origi- 

 nally from Ellwanger and Barry's Nursery under Loudon's name. 



