143 



actenstics in common. The difference of opinion on the part of 

 American botanists surprises him, as in Europe there has never 

 been any question in regard to the specific rank of the two forms. 

 Attention is then called to several points which merit careful ob- 

 servation. Thus R. humilis is stated to produce long roots, 

 which extend for some distance, producing suckers which event- 

 ually replace the parent plant. Earlier botanists do not seem to 

 have mentioned this characteristic, although Lindley remarks that 

 it is difficult to cultivate and especially to multiply, which was 

 probably due to the fact that these suckers were generally de- 

 stroyed and the original plants thus permitted to perish. It is no 

 doubt due to this cause that it has disappeared from cultivation. 

 Per contra, R. hicida, which is still common in cultivation, forms 

 permanent bushes, more or less tall and compact, increasing in 

 size year by year, and producing no stolons or suckers. It is for 

 American botanists to determine whether these points of differ- 

 ence hold good in the native plants, and if so, M. Crepin thinks 

 that it will be a strong argument for their specific rank. There 

 are also other points of difference in the inflorescence, prickles, 

 texture and autumnal coloring of the foliage, etc. Finally, the 

 careful study of the bract and sepals is urged, — the latter in re- 

 gard to position and appearance before and after maturity ; the 

 degree of decay of the petals, odor of the flowers and exact time 

 of flowering of the two species in any one locality. 



In regard to R. Carolina there seems to be no doubt in his 

 mind as to its specific rank, and he contends that the alleged in- 

 termediate forms, which Best says seem to connect with R. htinii- 

 lis, are most probably due to hybridization between the two 

 species ! He urges special study in this direction, as he cannot 

 imagine two species so entirely distinct being connected in any 



other 



way 



Rosa Arkansana is treated in a very summary manner and 

 the characters relied upon by our botanists to distinguish it from^ 

 R. hlanda are referred to ashaving but little value. Our author-' 

 ities are taxed with havinjr given to mere accident of growth a 



specific value, and also with emphasizing supposed points of dif- 

 ference which are in reality common to both species in many 

 cases. M. Crepin says thas wiiile he has striven to demonstrate 



