140 



REACTION-INTENSITIES OF STARCHES. 



WIDE EANGE OF REACTION-INTENSITIES. 



(Charts A 1 to A 20,) 



In comparing the range of reaction-intensities it 

 must be borne in mind that the values expressed in the 

 polarization, iodine, gentian-violet, safranin, tempera- 

 ture, and chemical-reagent charts are not formulated 

 upon the same basis of calibration. In the first four 

 instances the values are grossly quantitative, and the 

 abscissa? are founded upon crude and entirely arbitrary 

 standards and do not likely represent values that are 

 equivalent to those of the temperature or chemical-rea- 

 gent records. The temperature values are based upon 

 a scale that is different from those of the first group and 

 from those of the chemical reagents. The calibrations in 

 the first group, apart from the crudeness, are probably 

 defective because the reaction-intensities of the starches 

 studied do not extend, as in the case of those of the 

 chemical reagents, between the extreme limits of the 

 chart. The range in the temperature of gelatinization 

 charts closely resembles in its limitations the ranges in 

 the iodine, gentian-violet, and safranin charts. 



In these charts the abscissae-values, in comparison 

 with the corresponding values in the chemical-reagent 

 charts, are much too limited, but at present we have no 

 data which enable us to state (in terms of light, color, 

 and temperature reactions) the equivalent of a given 

 reaction-intensity that is expressed in time-per cent of 

 starch gelatinized. For instance, a difference of 2.5 

 in the temperature of gelatinization which is represented 

 by the space between two abscissae appears small on the 

 chart, yet this difference may have a differential value 

 that is equal to several times this abscissae-value in the 

 chemical-reagent charts. These temperature differences 

 would have been nearly equitably expressed in compari- 

 son with the chemical-reagent values had the tempera- 

 ture scale been between the extremes of say 50 and 85 

 instead of 40 and 95. A similar change could have been 

 made to advantage in the scales of the other charts men- 

 tioned. Comparing cursorily these five charts (A 1 

 to A 5), it will be noted that notwithstanding the com- 

 paratively limited ranges of reaction-activities each may 

 readily be distinguished from the others, with the excep- 

 tion of the gentian-violet and safranin charts, which are 

 very much alike and which, while easily differentiated 

 from the other charts, are distinguished from each other 

 only and doubtfully by careful comparison (see also 

 Chart B2). In fact, the differences in the latter are 

 unimportant because the crudeuess of the method of 

 valuation probably makes them fall within the limits 

 of error or observation. Among the chemical-reagent 

 charts the variations in reaction-intensities range in 

 nearly all, from reactions which are complete within a 

 few seconds to those in which so little as 2 per cent or 

 less of the starch is gelatinized in CO minutes. In ex- 

 ceptional charts (Charts A 10 and A 18, sulphuric acid 

 and sodium salicylate) the extent of the variations is 

 distinctly limited generally because of rapidity of gela- 

 tinization of the starches, in the former most of the reac- 

 tions being shown to be complete within 5 minutes, and 

 in the latter within 15 minutes. 



MANIFEST TENDENCY TO GROUPINGS OF EEACTION- 

 INTENSITIES. 



In both the preceding and present researches, par- 

 ticularly in the former because of the relatively large 

 numbers of species and varieties included among many 

 of the several genera, it has been found that the reaction- 

 intensities of the representatives of a genus tend to be 

 con lined usually within well-restricted limits, the max- 

 ima and minima reactions of members of the genus being 

 in general wider apart as they are botanically farther 

 separated, the greatest differences being noted when 

 specimens are included which belong to well-defined 

 generic subdivisions. Where the representatives of a 

 genus are not so far separated as to fall into such sub- 

 divisions, the variations tend to be confined to a space 

 on the charts that rarely exceeds 3 to 5 abscissae (22 

 being the chart limit), frequently less; but where there 

 are representatives that belong to different well-defined 

 subgeneric divisions (for instance, subgenera, tender and 

 hardy species, tuberous and rhizomatous forms, etc.) the 

 variations are, on the whole, much more extensive, 

 equivalent usually to the space of 10 to 20 abscissa? 

 or they may extend to practically the extremes of the 

 chart. As extraordinary as it may seem, while such ex- 

 treme variations may be found with one reagent, little 

 or no difference may be found with another reagent; 

 and with other reagents all intermediate values may be 

 noted between these extremes. These facts are well 

 illustrated in Begonia: No differences are noted in the 

 reaction-intensities of these starches in Charts A 10 

 and A 12 (sulphuric-acid and potassium-hydroxide reac- 

 tions), gelatinization in all being complete within less 

 than a minute; while in a number of other charts (as in 

 Chart A 9, the nitric-acid reactions) the same remark- 

 ably rapid reaction occurs in the starch of only one of 

 the parents and in the hybrid, while the reaction of the 

 other parental starch is remarkably slow. 



The extent of generic differentiation varies in the 

 different charts. Some differentiation is evident, for 

 instance, in Charts A 6, A 15, A 18 (chloral-hydrate, 

 potassium-sulphide, and sodium-salicylate reactions) ; 

 there is better differentiation in Chart A 7 (chromic- 

 acid reactions) ; and still better differentiation in Chart 

 A 8 (pyrogallic-acid reactions). The grouping of mem- 

 bers of a genus and the differentiation of the genus upon 

 the basis of reaction-intensities can be rendered satis- 

 factory only when large numbers of members of each 

 genus are studied ; when the maximum, minimum, and 

 average values are determined with a number of reagents ; 

 and when it is recognized that members of subgenera 

 and of other generic divisions may exhibit in the sum of 

 their reactions differences that may be as divergent as 

 those of different genera. For instance, in Nerine, it 

 will be seen that in 17 of the 26 charts the values of the 

 3 groups are within very restricted limits and constitute 

 a group of close values; and, moreover, that while the 

 maximum, minimum, and average values of the group 

 may be about the same as the corresponding values of 

 other generic groups, in certain reactions they will be 

 found to be different, so that in the final summing up 

 the genus stands very distinctly apart from the other 

 genera. In the remaining 9 charts there are varying 

 degrees of departure from this well-defined grouping, 



