M'MMAKlr> i'K 1IIK 1! I- I . >l.< ,\< c HAKA. IKIO. IK 



MM 



pollen parent ; in /ri< three incline to the need parent 

 and otic t<> the jollen parent; and in 1'iilanlke one in- 

 ilini-i t.. th.- Heed parent and one to the pollen parent 

 In the i/u<:'i!ilnlii-r reactions this absence of constancy 



or the other parent is uiuch more marked; thus, 

 in Miih /.Vufi-./..fi/i.i and lifijonia do all of these chloral- 

 hydrate ! to the seed parent ; but in no 



.- do all of thorn incline to the pollen parent. Exam- 

 ining the different generic groups we note that in Hip- 

 ptattrum in two h\hrids the reactions incline to the teed 

 parent and in one to the pollen parent ; in Harmanthiu 

 in one hybrid >n incline* to one an much u to 



iher parent, and in the other to the aeed parent; 

 -mum <>ne inclines to the need parent and two to the 



i parent ; in \rrinr one inclines to the feed parent 

 and fur t<> the pollen parent; in .\arcimnu five incline 

 to the seed parent, gix to the pollen parent, and two in- 

 cline to one a.- much aa to the other parent ; in l.ilium two 

 im-line to the *vd parent and three to the pnllon parent; 

 in Iris tn incline to one aa much u to the other parent, 

 and two incline to the pollen parent ; and in Calanthf 

 one incline* to the seed parent and the other inclines to 

 one as much as to the other parent. Of exceptional 

 interest is the fart, several times noted, that in case of 

 any hybrid the qualitative and quantitative reactions 

 may nr may not correspond in their parental inclinations. 

 ::!;. remarkable that with a given reagent the 

 qualitative reaction may correspond with that of the seed 

 parent and the quantitative reaction with that of the 



; parent, or rice versa, and so on in other varied 

 relationships. 



Th.- tendency in general to a ratio of approximately 



:i the qualitative reactions in their relations to the 

 seed and pollen parents is well marked. This ratio 

 varie* from 4 : to 1:1, but in about half of the cases it 

 will he found to be as first stated. Totaling these rec- 



it will be seen that 62.8 per cent of these reactions 

 incline to the seed parent and 35.8 per cent to the pollen 

 parent, a ratio of 1.8 : 1. In other words, there is 

 approximately twice the tendency for the qualitative 

 reaction to be closer to the seed parent than to the pollen 

 parent. 



There is not a corresponding tendency to such a com- 

 mon ratio in the quantitative reactions, but to a marked 

 inconstancy. In the qualitative reactions the ratio is 

 always in favor of the seed parent; but in the quantita- 

 -i-adions it may be in favor of either or of neither 

 parent. Thus, it is found that there may be a ratio 



I in favor of the seed parent, or one of 1 : 3 or 1 : 4 

 in favor of the pollen parent, and intermediate grada- 

 S itnming up these reactions, 44 per cent incline 

 to the seed parent and 40 per cent to the pollen parent 

 a ratio of approximately 1:1. In .-tudying the quanti- 

 tative records the large number of reactions that are 

 recorded as being the same as those of both parents 

 should be taken into consideration, because had these 

 been shown to have had in each ca.e, or even in most 

 cases, definite uniparental inclinations these ratios would 

 of course be subject to more or less modification. Nearly 

 all these reactions showed no difference from the parental 

 reactions because of gelatinization occurring with too 

 great a rapidity or slowness for differentiation. Modi- 

 fied strengths of reagents would doubtless have elicited 

 differences that are wholly obscured by very quick or 



<dow reactions. It i, however, not probable that there 

 uould be brought about any iui|>ortant change, aa a 

 whole, in these ratios. Why the qualitative ratios !. 

 be so different from the quantitative ratio* IK entirely 

 problematical, wry interesting, and very suggestive of 

 stereochemic peculiarities of the starches. 



feature of these records is more remarkable than 

 the reversal of the qualitative and quantitative reactions 

 of a given stan-h with a given reagent in their pm 

 inclinations. It is of importance to note that this phe- 

 nomenon is not peculiar to any starch or reageut, but is 

 common, and doubtless common to all starches and to all 

 reagents. With not a single starch was it found that 

 there was not such reversal ; and with onlv four of the 

 reagents (strontium nitrate, barium chloride, and mer- 

 curic chloride) was reversal not recorded, the rea-on for 

 which is doubtless to be found in the small number of 

 qualitative reactions recorded with these reagents (four 

 us with the first, one with the second, and four 

 with the third). Not lexs remarkable than tho reversal 

 of the reactions is the frequency with which this phe- 

 nomenon occurs, the percentages ranging from 6 in the 

 iodine reactions to as nigh as 50 in the cobalt-nit rate and 

 cupric-chloride reactions with the different starches. The 

 mean is 22.5, or close to one-fourth. 



TABU R. 



