162 



REACTION-INTENSITIES OF STARCHES. 





belladonna, and Brunsdonna sanderos (both the same), 

 Brunsvigia josephinoe, and Brunsdonna sanderw alba; 

 and in the sodium-sulphide chart, Brunsvigia josephince, 

 Amaryllis belladonna, Brunsdonna sanderce, and B. san- 

 derce alba. Viewing the various charts of this set, all 

 sorts of variations in the relative reaction-intensities of 

 these four starches will be found: In some, such as in 

 the charts for chromic acid, potaesium hydroxide, and 

 barium chloride, there are practically or absolutely no 

 differences; the charts for nitric acid, sulphuric acid, 

 and hydrochloric acid show some but not marked differ- 

 ences; the charts for chloral hydrate, potassium iodide, 

 potassium sulphocyanate, and cobalt nitrate show well- 

 defined pairing in all three reactions the parents and 

 the hybrids, respectively, are paired, in the chloral- 

 hydrate reaction the parental pair having the less reac- 

 tivity, while in the potassium-sulphocyanate and cobalt- 

 nitrate reactions the greater reactivity. In other in- 

 stances there may be a single pair, the other two starches 

 differing from this pair and from each other, as in the 

 reactions of pyrogallic acid, potassium sulphide, stron- 

 tium nitrate, cupric chloride, and mercuric chloride; in 

 other instances all four are unlike, as in the charts of 

 sodium hydroxide, sodium sulphide, calcium nitrate, 

 and so on. 



Pairing when present may be confined to either the 

 parents or the hybrids, or there may be pairing of both 

 parents and both hybrids, and in one instance (potas- 

 sium-sulphide chart) Amaryllis and Brunsdonna san- 

 derce are paired, and show distinctly different reaction- 

 intensities from those of the other parent and the other 

 hybrid, which two latter in turn differ markedly. In 

 other words, if any given set of parents and offspring be 

 taken and their reaction-intensities with the different 

 reagents be compared, it will be found that there are 

 not only very marked differences in the average reaction- 

 intensities of the several members with the different rea- 

 gents, but also most remarkable variations in the rela- 

 tive reaction-intensities with these reagents, so that 

 while a given starch may show the highest reactivity of 

 the set with one reagent it may show the least with 

 another, and so on, each starch being capable of reacting 

 in a way independently of the others, so that all possible 

 combinations of varying relationships may occur. This 

 means, of course, that in one reaction the hybrid may 

 be the same as that of the seed parent, in another the same 

 as that of the pollen parent, in another the same as the 

 reactions of both parents, in another intermediate, in 

 another in excess of those of either parents, etc. Each 

 reagent, therefore, has the property of eliciting some 

 definite parental phase. A somewhat detailed considera- 

 tion of this important phenomenon will be taken up in 

 Chapter V. 



VARIATIONS IN THE KEACTION-INTENSITIES AS EE- 

 OABDB HEIGHT, SUM, AND AVERAGE. 



(Table B 1, Chart C 1.) 



The valuations of the reaction-intensities have been 

 based, as has been repeatedly stated, on definite but arbi- 

 trary scales: Those of the reaction-intensities of the 

 polarization, iodine, gentian-violet, and safranin reac- 

 tions on a scale of to 105 ; those of the temperatures 

 of gelatinization on a scale of 40 to 95, and those of the 



reactions with the chemical reagents on a scale that shows 

 in one segment the percentage of total starch gelatinized 

 within 60 minutes, and in another the time of complete 

 or practically complete gelatinization within the same 

 period. Inasmuch as in all three sets the same abscissae 

 are used, and as the scale-values bear in all of the charts 

 the same relationships, the figures of one scale always 

 have a fixed value in relation to given figures of the other 

 scales; hence, if the scale for the polarization reactions 

 were adopted for valuation of all kinds of reactions the 

 values in all cases would be comparable upon a common 

 basis. For purposes of gross comparisons this scale has 

 been divided arbitrarily into 5 parts which are intended 

 to designate very high, high, moderate, low, and very 

 low reactivity, respectively. Thus, any reaction that falls 

 between 80 and 105 (or in the temperature scale 52.5 

 and 42.5 ; or in the chemical reagent scale 25 and 

 minutes), both inclusive, is recorded as being very high; 

 between 60 and less than 80, etc., as being high, etc. 

 Table B 1 gives, in connection with each starch, the num- 

 bers of the 26 reactions that fall under one or another 

 of these divisions; the sum of the individual reaction- 

 intensity values of each starch; and the average of this 

 sum, which latter is obtained by dividing by 2fi. Such 

 data constitute a very satisfactory basis for comparisons 

 of the reaction-intensities of the different starches indi- 

 vidually, generically, and so on, and they are rendered 

 of additional value if they are also reduced to chart 

 form. (Chart C 1.) 



The most conspicuous features of the table and chart 

 are: The close correspondence in the numerical distri- 

 bution of the reaction-intensities (very high, high, mod- 

 erate, low, very low) of the several starches of each set 

 of parents and hybrids and of each generic group, to- 

 gether with the close correspondence of the sum and the 

 average values, except when the set or genus represented 

 contains members of subgenera or subgeneric groups ; and 

 the varying values of the different generic groups. 



It will be seen, for example, in Hippeastrum, in which 

 generic group the parents are closely related, and where 

 consequently there is but little deviation in the reactions 

 of the hybrids from those of the parents, that the 

 figures in each of the columns of the chart for all of the 

 parents and hybrids are in close correspondence, and 

 that the sums and averages of the reaction-intensities are 

 also quite close. The range of these figures in the table 

 for all the starches studied is limited by 2614 (sum) 

 and 100 (average) in Cymbidium lowianum and 525 

 (sum) and 20 (average) in Hcemantlius katherina. Tn 

 the first column (very high reactivities) the figures range 

 from 2 to 4 ; in the second column, from to 3 ; in the 

 third column, from 3 to 5 ; in the fourth column, from 

 3 to 6; in the fifth column, from 11 to 14; in the sixth 

 column, from 748 to 925 ; and in the last column, from 

 29 to 36. These ranges will be found to be within very 

 narrow limits when compared with the figures of the 

 table, as a whole. Such correspondences are also well 

 marked in Nerine, Narcissus, Lilium, Oladiiolus, Tritonia, 

 Phaius, Miltonia, and Cymbidium. On the other hand, 

 when the genus is represented by bigeneric parents or by 

 members of subgenera or subgeneric groups, there may 

 be more or less marked deviations from those found when 

 the parents are monogeneric and not so far separated 



