148 



KKAf'TlON-IXTKXSJTIKS OF STA |;( II K>. 



B 12, it will be apparent thai a well-marked line 



of demarcation between these two groups; and 

 that when the five curves of Chart 1! 1 are com- 

 pared differences are exhibited that are in harmony 

 with the similarities and dissimilarities of the char- 

 acters of the reaction-processes. The polarization 

 curve stands in its peculiarities quite apart from 

 flie others, and it appears, on the whole, to be in 

 its course without more than incident.il relationship to 

 tie' courses of the other curves; hut the gentian-violel 

 and safranin cur\es show almost throughout their 

 -■■s, close correspondence in their variations with 

 each other (see also Chart B 2), yet an absence of corre- 

 spondence with the other three curves. Such differences 

 as are n corded in these two curves are doubtless attribu- 

 table to errors of experiment. When the crudity of the 

 method of valuation of these reactions is considered, it is 

 remarkable that the curves are so close, rather than that 

 there are some discrepancies. The iodine and tempera- 

 ture curves hear certain well-defined similarities, hut 

 they lack the close agreement seen in the two aniline 

 curves ; and they differ enough to indicate that the 

 -ses involved in the two reactions are not the same. 

 The absence of conformity of the aniline and iodine 

 curves, together with the agreement of the former, is 

 convincing evidence that here also the processes of the 

 two sets of reactions can not he the same. While the 

 iodine and temperature curves show similarities (Chart 

 B3) they differ as much in general from each other as 

 do the iodine and aniline curves. 



It will he seen that the iodine curve remains at vari- 

 distances above the temperature curve, excepting in 

 I. ilium tenuifolium, B. chalcedonicum, L. pardalinnm. 

 Iris iberica, Tritoma pottsii, and Phaius grandifblius, 

 where in 5 of the 6 it is below and in one the same. 

 The iodine valuations are only approximate, yet the 

 errors of observation are probably not sufficient to alter 

 the curve in any essential respect, at least in so far as 

 concerns general comparisons. On the other hand, the 

 temperature valuations are approximately scientifically 

 correct inasmuch as the errors of experiment fall within 

 such very narrow limits as not to affect apprecia b 

 the position of the curve at any point. While certain 

 variations in the quantitative differences between these 

 curves, and at points the inversioE and reversion of the 

 curves, might surest errors of valuation, they are in 

 conformity with the findings shown in the other charts, 

 as will be seen. Some of the variations of the iodine 

 U are probably due to differences in the behavior 

 of this reagent with the capsular and intracapsular parts 

 of the grains. NTageli found that iodine in weak solu- 

 tions may penetrate the capsular part to the intra- 

 capsular pari of the grains coloring the latter but not 

 the former. It would seem, therefore, that the iodine 

 of the raw starch grain-, as here studied, are 

 reaction e ntially, ami with weak solutions solely, of 

 the intracapsular part of the grain, and that the differ- 

 ences in color values of the reactions are -1 pi n lent in 

 part upon the peculiarities of the intracapsular starch, 

 and in part upon variations in the transmissive and 

 reactive properties With a given strength 



of iodine solution, when the grains are gelatinized by 

 heating, both intracapsular and capsular parts color, the 



former very much more than in the normal grain, and 

 the latter a different color from the intracapsular part — 

 the former blue, and the latter violet, old-ruse, etc. 



Eeating the starch grains in water, and various rear 

 gents gelatinize starch, but the molecular processes in- 

 volved can not, for reasons stated, be precisely the same. 

 The qualitative gelatinization changes in different 

 starches differ from each other; those caused by heat 

 differ from those caused by chemical reagents; and those 

 caused by one reagent differ from those caused by an- 

 other. The quantitative differences are in all corre- 

 sponding cases far more marked than the qualitative 

 changes. In the gelatinization caused by heat the change 

 in surface tension that gives rise to the inflow of wati r 

 is due, in accordance with our knowledge in general of 

 colloidal swelling, to ionic action. Both hydrogen and 

 hydroxyl ions are present, but it seems that the hydrogen 

 ion is the effective agent, and effective only at certain 

 temperatures that vary with the kind of starch. With 

 the chemical reagents there are not only hydrogen and 

 hydroxyl ions present, but also they are in compara- 

 tively very high concentration; and, moreover, there 

 are in the different solutions other kinds of ions and also 

 molecules that vary in kind and concentration. In these 

 reagents the ion concentration is without the aid of heat 

 sufficient to bring about the alteration in surface tension 

 that permits of hydration of the starch, and also there 

 are components of the solutions that with the ampho- 

 teric starch molecule may form various chemical com- 

 binations and influence the processes of gelatinization, 

 as previously stated. If these statements are justified, 

 such should be indicated when, for instance, the tem- 

 perature-reaction experiments are compared with those 

 of chloral hydrate, pyrogallic acid, nitric acid, and other 

 reagents. 



In comparing the curves of Charts B 4, B 5, and B 6, 

 it will be seen in each that the temperature-curve differs 

 markedly from the reagent curve, although there are 

 many suggestions of correspondence in the variations; 

 but they differ quite as distinctly from each other as do 

 the reagent-curves from each other. Moreover, not only 

 are there marked quantitative differences, but these dif- 

 ferences not infrequently take the form of inversion of 

 the curves, so that while with one starch temperature 

 vity may be higher than reagent activity, in an- 

 other starch there may he the reverse. For instance, in 

 the temperature chloral-hydrate chart (Chart B 4) it 

 will be seen that, here and there, varying direct and 

 inverse relationships in the up and down courses of the 

 curves occur, the one curve keeps continually above the 

 other with variable degrees of separation, and then the 

 curves will cross or become inverted, and at varying dis- 

 tances recross, such crossing and recrossing occurring a 

 number of times. Thus, the temperature curve is higher 

 than the chloral-hydrate curve in Amaryllis belladonna, 

 Hcemanihus leathering, 77. puniceus, Nerine bowdeni, 

 N. sarniensis var. corusca major, Lilium martagon, L. 

 tenuifolit n . L. chalcedonicum, h. pardaUnum, Iris fro- 

 jana, Bryonia single crimson scarlet, B. socotrana, and 

 Miltonia bleuana. In Amaryllis belladonna the tem- 

 perature curve is lower than the chloral-hydrate curve, 

 but in Brunsvigia Josephines the reverse. In the three 

 Ilippeastrums the temperature curve is the higher; the 



