A.MAUYU.IS liHl'NSVKilA- UKUNSIx >\ \ A. 



39 



Do the properties of Brunsdonna tuber g mi, Brunsdonna 

 sandera alba, and Brunsdonna sandera indicate that 

 these hybrids are the offspring of tin' same cross or of 

 reciprocal crosses; and (2) what are the indications 

 of tli.' probable parentage of Amaryllis parkeri alba'' 



Tin,' starch of Brunsdonna tubergeni has in compari 

 son with the starch of I:, sandera alba ami B. sandera 

 certain properties that arc closely similar or identical 

 and others that, are more or [ess markedly dissimilar, 

 the latter much predominating. The grains of the for- 

 mer are mure irregular, and more slender ami elongated; 

 the hila are less distinct; the lamellae are more distinct, 

 more often coarse, and more often irregular; the grains 

 are larger. In the polariscopic properties there are not 

 any conspicuous differences except that the figures tend to 

 be more irregular. In the iodine reactions the coloration 

 is distinctly less, In the aniline reactions with both 

 gentian violet and safranin the coloration is more 

 marked. In most of the foregoing instances the starch 

 of B. tubergeni does not differ more from the starches 

 of /.'. sandera alba and />'. sandera than do the latter 

 from each other. In the temperatures of gelatinization 

 the figure for B. tubergeni is 6-1.70°, or a difference 

 approximately of 7.5° less than the temperatures of 

 thi' parental starches, these being 72.'! ami ", l , re- 

 spectively. The temperatures for B. sandera alba and 

 B. sanderce are 72.25° and 72.75°, respectively. It will 

 be noted that while the temperature for the parental 

 starches differ only 1.7°, that of B. tubergeni differs 

 from that of the pollen parent {A. belladonna) 7.9-1°, 

 ami from that of the seed parent (/>'. josephina) (>.'.' 1 ; 

 and that the temperatures for B. sandera alba and II. 

 sandera and their parents differ very little, mostly within 

 the narow limits of error of experiment. The very low 

 temperature for B. tubergeni on the one hand and the 

 marked closeness of all of the temperatures for B. san- 

 dera alba and B. sandera: and their parents on the 

 other indicate quite conclusively that B. tubergeni and 

 B. sandera alba must have arisen from reciprocal crosses. 

 This conclusion is substantiated by the records (not- 

 withstanding their limitation) of the reactions with 

 chemical reagents. The reactions of all of the starches 

 with sulphuric acid occur with such rapidity that no 

 satisfactory differentiation is possible, but with both 

 potassium iodide and sodium hydroxide there are marked 

 and distinctly diagnostic differences. In reactions with 

 potassium iodide the starch of B. tubergeni exhibits a 

 somewhat higher reactivity than the starch of either 

 parent, while on the other hand the starches of B. sandt ra 

 alba and B. sandera -how very much lower reactivities, 

 not nearly so much of the latter being gelatinized at the 

 end of an hour as there is in case of the B. tubergeni 

 and parental starches in 5 minutes. It is also to be 

 noted that during the progress of gelatinization the 

 curves of B. sandt ra alba and />'. sandera tend to pursue 

 the same course, they being separated at and alter the 

 5-minute interval by about It) points. In the sodium 

 hydroxide reactions similar results are recorded, the 

 reactivity of the starch of 11. tubergeni being very high 

 and closely corresponding to the reactivities of the 

 parental starches, but slightly higher than either, while 

 the reactivities of the starches of II. sandera alba and 

 11. sandera are both moderate, the reactivity of the former 

 being distinctly lower than that of the latter. 



There were studied iu this research three group- of 

 parental ami hybrid - in each of which we 



d two hybrids of the same i ro , and it is of inter- 

 ■ Mo note to w in t di i general the membei - o 



pan- compare with each other and with their pa 

 and how I compare « ith those of the 



Brunsdonna hybrids and their parents. Examining 

 the temperatun latinization and taking up the 



Nerine en ins-dainty maid-queen of roses group 



i it w ill be seen that the tem 

 hi ids differ only 1.3° and that they are interim 

 between the parental temperatures, which Lai 

 5.2°; in the Nerint ! var. coi 



major giati e group the temperatu 



hybrids differ 3.35 and both are lower than of the 



parental temperatun 



Narcissus \ -poeticus poetarum-poeticu 



poeticus dante group the temperatures of the hj 

 differ 2 , that of one being n.i i the 



parental temperatures and the other practically the same 

 as that of the seed parent, while the parental tempera- 

 tures differ 5.5 ', thai of the seed p ing the higher. 

 The temperature- of eai h of th< e pairs of hyl 

 close together and close to the temperatures of the 

 parents, as in the case of Brunsdonna 

 B. sandcrw, with wider variations in the former than in 



the latter. bu1 there is no suggestion of a H I Hire, 



such as is found m H. tubergeni, this latter 

 either difference in parentage or in the direction i 

 ero-s from that of the other Brunsdonna. 



In the reactions of the members of these groups with 

 potassium iodide and sodium hydroxide corresponding 

 characteristics have been recorded, that is, that the two 

 starches of each group show eli 



In the potassium iodide reactions of the Nerine crispa- 

 elegans-dainiy maid-queen of roses group, those of the 

 hybrids are very much alike aid. on tic whole, inter- 

 mediate between Wi^-.- of the parents; and iii the Nerine 

 ni-sarniensis var. corusca major-giantess-abundance 

 group, while those of the hybrids are low and differ dis- 

 tinctly, at lea-t one and probably both tend to interme- 

 diateness, and one takes more after the seed parent and 

 the other more after the pollen parent. In the sodium- 

 hydroxide reactions, in the first group those of the hy- 

 brids are not only \ i\ close hut also clo-> to I 

 the parents ; ami in I d group those of t : , 



are very close and lower than those of the parents. It 

 will be seen that in the - of each of the se 



pairs of hybrids there are no such departures of the 

 reactions of each of the couples as an red in the 



i Brunsdonna tubergeni compared with B. sandera 

 ind II. sandera. from ; , description of 11. t\ 

 geni this hybrid is more closely related in its propi 

 to Brunsvigia josephina than to Amaryllis belladi 

 while the data of /;. sandera alia and B. sandera indi- 

 cate 'hat. on the whole, both of these hybrids sh 

 closer relationship to A. belladonna than to /;. j'>sejih- 

 ina — in other words, in each case the hybrid is more 

 closely related to the seed parent. 



These data also give a clue as to the probable or 

 of Amaryllis parheri alba. The starch of this plant 



i ;hout the In - ' and polariscopic prop. 



and thl md aniline reactions, with n ' ions, 



exhibits a i i Brunsdonna 



