142 



REACTION-INTENSITIES OF STARCHES. 



Nerine lines. In Chart A 15 (potassium-sulphide reac- 

 tions) Lilium remains the same; Nerine and Narcissus 

 are distinctly different, the lines of the former being 

 much shorter than those of the latter; and the lines of 

 Narcissus, Iris (all four groups), Gladiolus, and Tri- 

 tonia are all prolonged to about the same level, so that 

 there are no generic differentiations of these four genera. 

 In Chart A 18 (sodium-salicylate reactions) there is a 

 noticeable absence of resemblance of the lines collec- 

 tively to those of any of the preceding charts. Here, 

 Nerine, Narcissus, Lilium, and Iris (the first three sets 

 of the last) are, on the whole, very much alike. The third 

 set of Iris, which in the other charts shows greater reac- 

 tivity than the other three sets, now shows the opposite 

 relationship : and, moreover, while this set in the previous 

 charts is markedly different from Gladiolus and Tri- 

 tonia, here it is the same. Similar differences will be 

 found in other generic groups, in other sets, and also 

 with other reagents. These characteristics demonstrate 

 conclusively that the starches of different generic groups 

 and subgroups differ within wide limits in their molecular 

 structures ; that there are very definite generic and sub- 

 generic peculiarities ; and that these differences can satis- 

 factorily be reduced to figures and charts. 



INDIVIDUALITY OR SPECIFICITY OF EACH CHART. 



The individuality or specificity of each chart is very 

 pronounced and is most striking in the reactions in 

 which there occurs intermolecular disorganization of the 

 starch. Inasmuch as the starches are the same in each 

 of the charts (except in some instances as to number), 

 and the agents and reagents are variable, this individ- 

 uality is definitely associated with peculiarities of the 

 latter. Taking the charts, as a whole, it will be seen 

 that no two are alike, although in exceptional instances, 

 and for very obvious reasons, they differ in only minor 

 degrees and even within the limits of error of experi- 

 ment; well-marked examples of the latter are found in 

 the gentian-violet and safranin, and in the copper-nitrate 

 and cupric-chloride charts. On the other hand, where in 

 accordance with general laboratory experience no mate- 

 rial differences should be expected, excepting such as 

 would be dependent upon differences in the concentra- 

 tion of the reagents, as in the potassium and sodium- 

 hydroxide charts, respectively, the individualization is 

 not only very marked, but also in a measure entirely 

 independent of differences in concentration. 



As previously stated, these 26 charts fall naturally 

 into two primary divisions in accordance with whether or 

 not in the reactions there occurs intermolecular disor- 

 ganization. In conformity with recognized principles of 

 physical chemistry, comparatively limited variations 

 should, as a rule, be expected when in the reactions the 

 starch molecules remain wholly or apparently intact, as 

 in the polarization, iodine, gentian-violet, and safranin 

 reactions; but wide to extremely wide variations when 

 the molecules are broken down, especially in cases of 

 reagents which may have multiple active components 

 taking part in the disintegrative processes. As previously 

 stated, the polarization reaction is a light reaction in 

 which the moleciiles are undisturbed ; the gentian-violet 

 and safranin reactions are, in all likelihood, adsorptivc 

 phenomena which, as far as known, do not involve dis- 



arrangement of the starch molecules ; and the iodine reac- 

 tion seems to be of a kind in which an unstable iodide 

 of starch is formed, but without obvious intermolecular 

 disorganization ; the temperature reaction is one of hy- 

 clration which, while causing intermolecular breaking 

 down, does not give rise to a loss of typical starch proper- 

 ties; and the reactions with the various chemical rea- 

 gents are primarily phenomena of hydration, such as are 

 brought about by heat, but modified quantitatively and 

 qualitatively by differences in the components of the 

 reagents which take part in the reaction. 



It is obvious that the polarization reactions stand 

 entirely apart from all others ; that the gentian-violet and 

 safranin reactions constitute an isolated pair ; that the 

 iodine reactions stand by themselves; and that the tem- 

 perature and chemical-reagent reactions form a well- 

 defined group, the former representing one and the latter 

 another subgroup. In the temperature reaction we have 

 a typical manifestation of the simplest form of the proc- 

 ess of gelatinization, while in the chemical-reagent sub- 

 group there is this same type but which is more or less 

 materially modified by various substances that have 

 chemical relations to the starch molecule. A comparison 

 of the temperature and chemical-reagent charts will show 

 that the latter not only differ markedly from the former, 

 but also as much or more from each other. It would 

 seem to follow, as a corollary, that the more varied and 

 widespread the chemical disturbances in the starch mole- 

 cules the more varied the reactions and the better the 

 differentiation of genera, species, parents, and hybrids. 



The individuality of each of the chemical-reagent 

 charts that is definitely associated with peculiarities of 

 the reagent is due in part to concentration and in part to 

 composition of the reagent. This salient point is elicited 

 clearly when the data recorded in any two arbitrarily 

 selected charts are compared. Thus, taking Charts A 6 

 and A 7 (chloral-hydrate and chromic-acid reactions) 

 a first glance will indicate that the average length of 

 the ordinate in the former is greater than in the latter 

 and, hence, that the concentration (reactive-intensity 

 of the reagent) is less than in the latter ; but it will also 

 be very apparent, upon comparing the lengths of the 

 ordinates of any given set of parents and hybrid, or of 

 any generic group in the two charts, that the differences 

 are . not such as are to be expected were the reaction- 

 intensities exhibited by those reagents dependent solely 

 upon differences in concentration. 



Should the differences in the reaction-intensities de- 

 pend merely upon differences in concentration (as of the 

 same reagent) it seems obvious that if with a given starch 

 the reaction with one reagent is equal to the length of 

 say 2 abscissas, and with another reagent to the length 

 of 3 abscissas, a corresponding though not necessarily 

 proportional relationship should be found in the reactions 

 of the different starches. In fact, not only may there 

 be an entire absence of such quantitative relationship, 

 but also a reversal of reaction-intensities, the reagent of 

 higher concentration being the stronger in some reactions 

 but the weaker in others. Thus, in Chart A 6 (chloral- 

 hydrate reactions), in the Amaryllis-Brunsvigia-Bruns- 

 donna set, it will be seen that the ordinates for Amaryllis 

 and Brunsvigia extend to the abscissas values 96 and 82, 

 respectively, and that those for the hybrids extend to 



