REACTION-INTENSITIES WITH EACH AGENT AND REAGENT. 



159 



distinctly differentiated by the inversion or reversion 

 of the reactive-intensities of a given pair of reagents, 

 but not by another pair. Thus, in the chloral-hydrate 

 and nitric-acid reactions (Chart B 11) the first inversion 

 seen occurs in the curves between Hippeastrum and 

 Hcemanthus, the three species Hcemanthus showing a 

 higher reactivity with nitric acid than with chloral hy- 

 drate, while Hcemanthus katherince shows the reverse. 

 But the differentiation here is not generic because the 

 second species, Hirmanthus puniceus, exhibits a reversion 

 in relation to the first species. In the chromic-acid 

 and pyrogallic-acid reactions the reverse is noted in 

 the behavior of these two species, H. katherince showing 

 in common with Hippeastrum a higher reactivity with 

 chromic acid, while H. puniceus shows the inversion. 

 In other charts (as, for instance, in Chart B 32 and 

 B 3G) all species of Hippeastrum and Hcemanthus show 

 in common a higher reactivity with one of the two rea- 

 gents.; while in other charts there are various modifica- 

 tions. For instance, in Chart B 35 each Hippeastrum 

 shows different reactivities with the two reagents, but the 

 Haemanthuses no difference. 



Crossing of the curves occurs again between Nerine 

 bowdeni and N. sarniensis corusca major, thus markedly 

 differentiating the first from the last two species of this 

 generic group. The same separation will be seen in 

 Chart B 2 (gentian violet and safranin), while in Chart 

 B4 (chloral hydrate and temperature) and Chart 8 (ni- 

 tric acid and iodine) the crossing occurs between N. 

 crispa and N. bowdeni. The next crossing occurs between 

 7m and Gladiolus; the next between Tritonia and Be- 

 gonia and the next between Begonia and Phaius all rep- 

 resenting generic lines of division. Comparing the 

 locations of these points of inversion or reversion with 

 those in the nitric-acid and chromic-acid chart (Chart 

 B 12) it will be found that with two exceptions (between 

 Iris and Gladiolus, and between Tritonia and Begonia) 

 the points are entirely different. The first crossing here 

 occurs between Brnnsvigia and Hippeastrum; the second 

 between Hcemanthus and Crinum; the third between 

 Crinum moorei and C. zeylanicum; the fourth between 

 C. zeylanicum and C. longifolium; the fifth between Ne- 

 rine sarniensis var. corusca major and Narcissus; the 

 sixth between Narcissus and Lilium; the seventh between 

 Lilium and Iris; the eighth between Iris cengialti and 

 7. persica var. purpurea : the ninth between 7m and Glad- 

 iolus; and the tenth between Tritonia and Begonia. 

 Some of these ten inversions and reversions occur between 

 generic representatives, while others represent subgeneric 

 dividing lines. 



The different points of inversion and reversion of the 

 curves shown in these charts (Charts B 1 to B -10) are 

 exhibited collectively in Chart B 41, this presentation 

 rendering further detailed statement in regard to each 

 chart unnecessary. Even a superficial study of the vary- 

 ing points of crossing of the curves and of the totals of 

 this chart brings out very interesting and significant com- 

 parisons. In confirmation of statements made in preced- 

 ing pages, it will be found that in some of the charts (12 

 out of the 40) no crossing of the curves occurs at any 

 part ; that in most of the charts there are inversions and 

 reversions, the number ranging from 3 to 10; that inver- 

 sions and reversions are, on the whole, more common 



when the agents and reagents are of dissimilar character 

 and when they exhibit wide and frequently varying 

 ranges of reaction-intensities; and that the crossings 

 of the curves are most apt to occur at points of separation 

 of genera and subgeneric representatives, and in variable 

 numbers with different reagents and different starches at 

 such places. The closely related genera Amaryllis and 

 Brunsvigia are distinguished by the inversion of the 

 reactions in only a single instance ( Chart B 4, tempera- 

 ture and chloral-hydrate reactions). Brunsvigia and 

 Hippeastrum have a separation by 9 crossings, but the 

 latter is separated from Hcemanthus by only 3. Curi- 

 ously, the two species of Hcemantlius are separated by 6 

 crossings, these variations of the curves suggesting sub- 

 generic division of the species. Hcemanthus is separated 

 from Crinum by 8 crossings, and Crinum from Nerine 

 by 7 ; but there are 9 between Crinum moorei and C. zey- 

 lanicum, and 11 between the latter and C. longifolium, 

 markedly differentiating the two hardy forms from the 

 tender form. The separation of Nerine from Crinum and 

 from Narcissus is well marked, there being 7 crossings 

 at the former point and 14 at the latter. Narcissus is 

 separated from Lilium by 9, and the latter from Iris by 

 15. The separation of the first three I rids from the 

 fourth is evident by 8. Gladiolus and Tritonia, are 

 separated by only 3, but these two are separated from 

 Iris by 12 and from Begonia by 11. The remarkable 

 differences exhibited by the tuberous and semituberous 

 Begonias are here illustrated by the separation of the 

 two by 15 crossings. Begonia is separated from Phaius 

 by 7, and Pliaius from Miltonia by 8. 



Wide Differences in the Reactions with Different 

 Pairs of Reagents. Another feature of exceptional in- 

 terest is the wide differences in the reactions of different 

 pairs of starches with different reagents, as has been 

 referred to repeatedly, and which is worthy of some 

 special notice. This peculiarity is well exemplified, for 

 instance, in Amaryllis and Brunsvigia. Little or, in 

 some instances, no difference is observed in the 

 reactions of these starches with chromic acid, sul- 

 phuric acid, hydrochloric acid, nitric acid, potas- 

 sium hydroxide, potassium iodide, potassium sulphocya- 

 nate, sodium sulphide, cobalt nitrate and barium chlo- 

 ride; distinct but not marked differences are noted with 

 chloral hydrate and sodium salicylate; and marked dif- 

 ferences are recorded with pyrogallic acid, potassium 

 sulphide, sodium hydroxide, calcium nitrate, uranium 

 nitrate, strontium nitrate, copper nitrate, and cupric 

 chloride. The reactions of Amaryllis are higher than 

 those of Brunsvigia. with chloral hydrate, nitric 

 acid, hydrochloric acid, sulphuric acid, potassium sul- 

 phide, sodium hydroxide, sodium salicylate, calcium ni- 

 trate, uranium nitrate, strontium nitrate, cobalt nitrate, 

 and cupric chloride; lower with pyrogallic acid, potas- 

 sium hydroxide, potassium iodide, potassium sulphocya- 

 nate, barium chloride, and mercuric chloride; and the 

 same with chromic acid and sodium sulphide. Even 

 better illustrations are to be found with other pairs of 

 starches, as, for instance, the two Begonias. 



Limitation of Number of Gelatinizing Reagents, Et'c. 

 The variety of the reagents used in this research to 

 gelatinize starch, together with the amphoteric proper- 

 ties of the starch molecules, may give the impression 



