Kl \< nnN-lNTKXSITlKS \\HII I. MM .M.I.M AND III M, INI 



\ -.h may exhibit like or unlike reaction! 



with ilitfi-r.-nt reagents, and the curvi-s vary a* much as 



.I.. those of d; ' .r, h,- with the same reagent, to 



that there may be moat varied forms of t .- liinYrent 



11ns feature will be found to be well exhibited 



hi-n the .ur\c> of tin- reactions of any given star 



any une of the generic groups are com. r in- 



.-, the i urvea of Amaryllis brllatlonna (Chart l> 1 t" 



ur\e 111 the chloral-hydrate reaction it of 



fiirin, having mi iiieliimtKUi of about 50, to that 



:|i|M-r en.l is at the terminutii'ii of the CD-minute 



interval. The eurve of the chromic-a. ill react imi 



the f form, hut it terminate* at the end of the 30-ininutv 

 il. -^.Mg it an inclination of about 30, which 

 . mil. h more rapid gelatinization. It will 

 be seen, h .< . r. that during the tint 5 minutis the 

 gelatinized in both reacti"ii- i- practically 

 the tame (1- and 10 per cent, res|>e -lively i, that th- 

 .11 the . :. 1 reaction occurs during the next 



HI minutes; ami that the quantities gelatini/.eil during 

 the interval between !." and 30 minute- are the same in 

 both reactions. The pyrogallic-acid and ehloral-hydrate 

 s bear a clone n-*-nil)laiur ; hut the former is lower 

 throughout, especially at the end of the 5-minute inter- 

 val. indicating a more marked early resistance t> thin 

 reagent than to chloral hydrate. From tin- point on- 

 ward to the end of CO minutes the curves run very closely 

 pa rail. 1. 



In 11 of the 21 experiments with different reagents 

 irves belong to the form of circumlinear type that 

 . rized by progressively decreasing increments of 

 starch gelatinized during additional increments of time. 

 These carves vary markedly in character. In some the 

 .'iient of starch gelatinized during the first 5 minutes 

 ry disproportionate to the quantities subsequently 

 broken down, as is noted particularly in the reactions of 

 potassium sulphide, sodium hydroxide, calcium nitrate, 

 -trontium nitrate (Charts D 10, D 11, D 14, and 

 lli. i. in each of which about 98 per cent of the total 

 starch was gelatinized in 5 minutes. In the sodium- 

 mlp n.- the increments of gelatinized starch 



an- (if,, 1 1. i. :t, und per cent. In the other reactions 

 f tin- group, in. hiding those of potassium iodide, so- 

 dium salicvlate. uranium nitrate, copper nitrate, and 

 chloride (Charts D8, D 13, D 15. D 18, and 

 i. tl.e curves exhibit various modifications in com- 

 parison with the foregoing. In the mercuric-chloride 

 reactions the curve is of a modified / form, tending, in 

 fact, like the accompanying Hnmtriyia jottphintr curve, 

 rectilinear, but at an angle of about 18 as com- 

 pared with about 26 for the latter. In the reactions of 

 nitrii m id. -ulphuric acid, hydrochloric acid, and potas- 

 v (Charts D4, D5, DC. and I) 7), th. 

 linear and almost vertical, while in the 

 barii; !e reactions (Chart D20) it is rectilinear 



and almost horizontal. 



irches of members of a genus tend, as a rule, in 



their reactions with each reagent to yield curves that are 



:.. l.i..- to the name type and type form, except when 



are snbgeneric representatives or widely separated 



. which case it may be found that there is or is 



not relationship in the characters of the curves, and this 



peculiarity may also apply to the curves of hyhr 



relation to those of its parents. For instance, taking 



the chloral-hydrate reactions : of the starches of LUitim 

 (Chart* l..i:. 1' .. I' '...and 1>373) the concord- 

 ance of both type and tvpe-form is obvious; of the 

 starches of Xtriiu (Charts D 190, 1)211, and D'j;).'), 

 the curves of the five parental starches are of the / form, 

 but vary in their courses >uili. i, ntly for easy differentia- 

 tion; of the starches of Crinum Mimrri. <'. Innyifiilium 

 and C. potrellii compared with those of ('. try/an irum, 

 where we have suhgeneric or the equivalent of aubgenerio 

 : jT.vntatives (Chart* D K 1 :. D IIS, and D Hi-.M. the 

 i ur\en of the first thnv c.niform to a given type-form, 

 while the curve of the latter is of an entirely different 

 type; of the starches of Hegonia, where similarly well- 

 neparatod starches are represented by those of the aeed 

 parent on the one hand and by the starch of It. socolrano 

 (pollen parent) on the other (Charts I) H,:<, l> 

 1)533, and D539), the curves are closely similar; of 

 the starches of Amaryllu and lintnurigia, where two 

 recognized genera are represented, the curves arc imii h 

 alike (Chart Dl). Varieties that are olTsjirin 

 closely related parental stock, as in Hippeaxlruin (Charts 

 I > -'-'. I > 43, and D 64), tend to show marked closeness in 

 the curves and this may also be seen not only in closely 

 related species, as in J'hain* (Chart D. r iTl) and Irit 

 (Chart D 4'.'1 ), hut also in closely related penera, as in 

 (Hadiolu* and Trifonia (Charts I) |f,:i and 1> |s| ). The 

 .iir\es of hybrids show, as will be pointed out particu- 

 larly hereafter, the most varied relationships to the 

 parental curves, varying between identity and great 

 dissimilarity. 



Taking the reactions of all of the parental starches 

 with any given reagent and comparing them with those 

 of other reagents, it becomes apparent that those of each 

 reagent represent a group in which there are both simi- 

 larities and dissimilarities ; and that the different groups 

 as such exhibit similarities and dis-imilantie-. the reac- 

 tions collectively of each group iM'ing quite as or even 

 more distinct from those of another group as are those 

 of members of the same group; that the more closely 

 related the starches the more marked the tendency . 

 ally to closeness of the curves, yet sometime^ distantly 

 or wholly unrelated starches may exhibit almost if not 

 identical curves with a given reagent. In a word, the 

 {x-culiarities of these reactions are of such characters 

 as should logically be expected if we are dealing with 

 stereoisomeric forms of staroh. 



The starches of the hybrid and parents usually take on 

 within a brief period after the beginning of gelatinization 

 definite relationships, which may be the same or different 

 in the reactions with different reagents. That is, if 

 shortly after the beginning of the reaction the |M.sitions 

 of the three carves should he in the order of intensity 

 of reactivity, seed parent, pollen parent, and hyhrid ( high- 

 . -t. intermediate, and lowest), this relationship usually 

 tends to be continued during the entire period of gela- 

 tinization, but with varying degrees of separation of the 

 curves. The hybrid curve may bear any relationship 

 to one or the other or both parental curves that is, be 

 higher or lower than either, or intermediate, or the same 

 as one or the other or both. Itarelv the parental curves 

 crow (Chart D169), or the hyhrid curve crowes one 

 or the other parental curve (Chart 089). The hybrid 

 curves tend usually to follow closely the parental curves, 

 but they may differ as much or more from the parental 



