SUMMARIES OF THE HISTOLOGIC CHARACTERS, ETC. 



287 



hybrid to Amaryllis belladonna, in the copper-nitrate and 

 cupric-chloride reactions it is not so close as the other 

 hybrid. 



HippEASTBUM. (Table C 2.) 



In comparing these records and keeping in view the 

 botanical closeness of the parents in each case, and also 

 a corresponding closeness of the offspring to the parents, 

 together with the great importance that is commonly 

 attached to intermediateness as a criterion of hybrids, 

 one is struck by (1) the frequency of the development 

 of properties of the hybrid in excess or deficit of parental 

 extremes; (2) the appearance of reactions in the hybrid 

 which were not seen in the parents ; and (3) the swinging 

 of hybrid development to one or the other parent in an 

 utterly inexplicable manner. Among the 36 designations 

 of the three sets, in no less than 23 some property or 

 properties were developed in excess of parental extremes, 

 and in 4 there was deficient development. In two in- 

 stances properties were noted in the hybrid that were 

 not apparent in either parent. The hybrid of the first 

 set is in form closer to the seed parent, but in the second 

 and third sets it is closer to the pollen parent ; in hilum, 

 in the first and third sets, closer to the seed parent, but 

 in the second set closer to the pollen parent ; in lameUse, 

 in the first set closer to the pollen parent, in the third 

 set closer to the seed parent, and in the second set closer 

 to the seed parent in number and to the pollen parent in 

 general characters ; in size, in the first set closer to the 

 pollen parent, in the second set closer to the seed parent, 

 and in the third set equally like both parents in common 

 size, but like the pollen parent in the larger grains. 

 In polariscopie figures and reactions with selenite, in the 



first and second sets the hybrids are more like the seed 

 parent, but in the third set the likeness is to the pollen 

 parent. The qualitative reactions with the chemical 

 reagents are full of interest. In the first set, with all 

 five reagents the reactions are, on the whole, closer to 

 those of the seed parent; in the second set those of three, 

 of the reagents (chloral hydrate, potassium iodide, and 

 potassium sulphoeyanate) are closer to those of the 

 seed parent, and two (nitric acid and sodium salicylate) 

 closer to those of the pollen parent ; and in the third set 

 those of four of the reagents are closer to the seed parent 

 and that of one (sodium salicylate) as close to that 

 of one as to that of the other parent. The relationships, 

 on the whole, are somewhat closer to the seed parent. 

 The quantitative and qualitative reactions show com- 

 paratively the most variable relationships. 



H^MANTHUs. (Table C 3.) 



The hybrid in the first set, in form and hilum, is 

 closer to the seed parent; in lamellae it resembles both 

 parents in equal degree; and in size it is nearer the 

 pollen parent. In the second set, in aU four histologic 

 designations, it is nearer the pollen parent. In the 

 polariscopie figures and selenite reactions and in the 

 qualitative reactions with the chemical reagents the 

 resemblance (except the iodine reaction in the second 

 set) is closer to the seed parent. In three instances 

 development in excess of parental extremes, and in one 

 instance individuality, were recorded. The quantitative 

 reactions are most vagarious in their relations to the 

 qualitative reactions. It is of interest to note that the 

 seed parent is the same in both sets and that in both 



Table C3, — HcBmanihus. 



