KKACTION-1M KVM 1'IES WITH BACH AGENT AND REAGENT. 



151 



reactivities with chromic acid are moderate and those 

 with pyrogallic acid very Low; while tlie corresponding 

 reactivities with 11. puniceus are high and very high, 

 respectively. The chromic-acid reaction is as much 

 higher than the pyrogallic-acid reaction in //. Jcatherina 

 as n i- lower in 11. puniceus. This interesting inver 

 sion of reactive intensities of the two starches « ith these 

 reagents is consistent with well-separated characters of 

 spei ies, as already pointi i out. Jn Crinum the two 

 hard) species are much more reactive to chromic acid 

 than to pyrogallic acid, whereas the reverse relationship 

 is seen in the reactions of the tender species; moreover, 

 curves of the latter are inverted in comparison with the 

 former. In Nerine the chromic-acid reactions are mod- 

 erate, while those of pyrogallic acid are so very low as to 

 he almost absolutely negligible, making a very marked 

 difference between the reaction-intensities. In Nari 

 the chromic-acid reaction is moderate and the pyrogallic- 

 acid reaction low, but without much difference between 

 them. In Lilium all of the reactions are high to very high, 

 the chromic-acid reactions being the higher except in one 

 species, in which both reactions are the same, although 

 during the earlier part of the experiments chromic acid 

 showed a somewhat higher reactive intensity than 

 pyrogallic acid. 



The degree of separation of the two curves in the 

 other three specimens is not alike in any two. In Iris 

 the chromic-acid reactions are high in all four starches, 

 and the pyrogallic-acid reactions moderate in two, low 

 in one, and very high in one. The distance between the 

 curves is marked in all four, and in /. persira var. 

 purpurea the curves arc inverted — in other words, the 

 lirst three starches are more sensitive to chromic acid than 

 to pyrogallic acid, while in the last there is the reverse. 

 Throughout this group of charts it will be seen that this 

 form uf Iris exhibits a number of peculiarities of reac- 

 tivity which definitely differentiate it from the preceding 

 three, which in turn seem to be closely related in 

 their reactivities. Inversion and reversion of the curves 

 of the irids corresponding to the foregoing will be found 

 in Charts B 7, B 8, B 9, B 10, B 12, B 22, and B 36. In 

 Gladiolus and Tritonia the chromic-acid reactions are 

 high and the pyrogallic-acid reactions moderate, the 

 reactions of the two starches with each reagent being 

 the same or practically the same, but the reaction-intensi- 

 ties with the two reagents being markedly different. In 

 Begonia the chromic-acid and pyrogallic-acid reactions 

 are distinctly higher in Begonia single crimson scarlet 

 than in B. socotrana, and the difference between the two 

 reactions is very much greater in the latter than in the 

 former. In Phaius and Mil Ionia the chromic-acid reac- 

 tions are much higher than the pyrogallic-acid reactions, 

 but the amount of separation between the two curves is 

 nearly the same. 



Examining this chart (B31) from the aspect of 

 generic and subgeneric differentiation, it is essential 

 to bear in mind that certain genera are represented by 

 individuals that show such marked differences as to 

 indicate that they belong to subgenera or some other 

 form of subgeneric division, as in Hcemanthus, Crinum, 

 Iris, and Begonia, and that on this account variations of 

 their curves may be such as to appear to be oppose! to 

 recognized generic grouping. With this peculiarity in 



view, beginning with Amaryllis and Brunsvig\ 

 related genera), it will be seen the p 

 curves in each arc very different -in Amaryllis the two 

 - are well separated, but in Br i they are 



the same. There is hi re ad f the two 



genera. Th< e genera are well separated from Hi. 

 trum, and the latter from the Hcemanthus, by the marked 

 differences in the i urvi . I [orms ol 



peastrum the chromic-acid curve is higher or even much 

 higher than in 1 ding and enera, and 



it i- in two well above and in one d e the 



pyrogallic-acid curve. I . the 



group- are so different that 

 one could not possibl) be confounded with another. In 

 Hcemanthus there is a drop of the chromic-acid curve 

 in 11. Jcatherina and //. puniceus; and a very ma 

 drop of the pyrogallic-acid curve in the former, 

 marked rise in the latter, giving rise to a well-de 

 separation of this genus from Hippeastrum and to inver- 

 sion of the curves in //. puniceus with consequent separa- 

 tion of the two species. In Crinum the picture is again 

 different, there being a rise of the acid curve 



accompanied by a rise of t allic-aeid curve in 



two and a fall in one. 



Enversion of the curves occurs in relation to ('. zey- 

 lanicum, this feature of itself differentiating this ; 

 species from the two hardy spe - In Serine the pic- 

 ture is again and markedly altered. Both curves fall, 

 the chromic-acid curve to a mode; I and the pyro- 



gallic-acid curve almost to zero, and with very little or 

 practically no difference in the reactivities of the four 

 starches with each of the reagents. In Narcissus, while 

 the chromic-acid curve remains at practically the same 

 level as in Nerine the pyrogallic-acid curve has 

 almost to the level of moderate reactivity, thus causing 

 some separation of the two curves and giving a generic 

 combination of the two curves which differs from that 

 found in any other part of the chart. In Lilium the 

 picture is again changed and is again distinctive of the 

 genus. And so on, as we pa-- to Iris, Gladiolus and 

 Tritonia, Begonia. Phaius, and .1/ the curves \ary 



in their positions and degree of separation in such man- 

 ners as to differentiate or suggest, a- the case may be, 

 not only generic hut subgeneric groups. The Gladiolus 

 and Tritonia curves are practically identical, the exp 

 tion for which lias been referred to repeatedly. The 

 first three and the last of the Iris are well separa 

 but Begonia shows curves of the two starches which, 

 while well separated, rather indicate well-separated spe- 

 cies than representatives of subgenera, as in the case of 

 many of the other charts. 



While it is true that in a number of instances a genus 

 is represented by only a single species and that, inasmuch 

 as tie- reactivities of different of a genus exhibit 



varying reactivities with the same reagents and thus 

 g;esi that the differences (in so far as they are applied 

 to tie- differentiation of genera) may he merely casual, 

 it will nevertheless he f clear by examina- 



tion of the accompanying charts that the evidence in 

 port of the generic and subgeneric differential 

 other relations here noted is cumulative and convincin r . 

 The very marked differences in the reactivities of Bub- 

 generic groups which are quite as rreat. on the wl 



