40 JOURNAL OF THE ARNOLD ARBORETUM [vol. hi 



1797, though a tree growing in the garden of the Bishop of London at 

 Fulham which Miller refers to his R. echinata and describes as having 

 much shorter pods densely beset with short prickles but otherwise re- 

 sembling R. pseudo-acacia may have been R. viscosa; the fact that this 

 tree " produced plenty of seeds" also speaks against its hybrid nature. 

 Even if the identity of the Fulham tree should be established, the name 

 R. echinata could not be applied to it as it is based primarily on Boer- 

 haave's Pseudoacacia vulgaris' which is partly Robinia pseudoacacia (as to 

 Tournefort's name), while the other citations refer to a tree figured by 

 Cornut (Canad. PL Hist. 172) which has subsessile flowers and one-seeded 

 spiny legumes and is not a Robinia. 



The adoption of the name R. ambigua makes necessary the following 

 new combination. 



Robinia ambigua var. bella-rosea, comb. nov. — R. bella-rosea Nichol- 

 son, Diet. Card. in. 310 (1887).— Mottet, Diet. Ilort. iv. 515 (1890-7). 

 R. viscosa f. bella-rosea Voss, Vilmorin's Blumengart. i. 219 (1894). — 

 R. dubia var. amoena Hort. apud Mouillefert Arb. Arbriss. I. 567 (1894). — 

 R. pseudo-acacia var. bella-rosea Cowell in Bailey, Cycl. Am. Ilort. 

 iv. 1538 (1902). — R. dubia var. bella-rosea Rehder in Mitteil. Deutseh. 

 Dendr. Ges. xxiv. 223 (1915). 



This form is nearer to R. viscosa while the typical R. dubia is closer 

 to R. pseudoacacia. When and where this form originated I have not 

 been able to find out, but it must have originated before 1880, for in 1883 

 it was in cultivation at Kew, as specimens in the herbarium of the Arnold 

 Arboretum collected in that year by G. Nicholson show. If R. viscosa 



bellidiflora Hort. apud Eismann in Hamburg. Gart. Blumenzeit xxxiv. 

 117 (1878) belongs here, is doubtful. 



Caragana frutex K. Koch var. macrantha, nom. nov. — C. frutescens 

 var. grandiflora Rehder in Bailey, Cycl. Am. Hort. I. 242 (1900), not 

 Regel (1866). — C. frutex var. grandiflora Koehne Herb. Dendr. no. 514 

 (1904), in sched. — Schneider, 111. Handb. Laubholzk. n. 103, fig. 64w-y 

 (1907).— Komarov in Act. Hort. Petrop. xxix. 226 (1908).— Rehder in 

 Bailey, Stand. Cycl. Hort. n. 160 (1914). 



As the name C. frutescens var. grandiflora Rehder is preoccupied bj 



C. frutescens var. grandiflora Regel in Bull. Soc. Nat. Mosc. xxxix. 570 

 (1866) it is not a valid varietal name and must be changed even though 

 Regel's varietal name, which is based on C. grandiflora DC. is generally 

 referred as a synonym to that species which is kept distinct by most au- 

 thors or referred to C. pygmaea as a variety. Caragana frutex var. ma 

 crantha occurs apparently occasionally with the type for one branch of a 

 specimens collected by P. N. Krilof near the village Lokot, western Si- 

 beria, on May 14-16, 1901, represents this form, while the other is typical 

 C. frutex. 



