340 



SUMMARIES OF PLANT CHARACTERS, ETC. 



BO remarkable among these records as the marked ten- 

 dencies in the several sets of parents and hybrids to 

 inverted relations of macroscopic and microscopic values ; 

 and the tendency for macroscopic values to be higher 

 than the microscopic values in the intermediate charac- 

 ters, and for the reverse in the characters that are de- 

 veloped in excess and deficit of parental extremes. 



Recapitulating the sums of both macroscopic and 

 microscopic characters that fall under the six sex-phases 

 (Table I, Summary 1) it is found that of the 959 charac- 

 ters 5.8 per cent are the same as those of the seed parent, 

 6.8 the same as those of the pollen parent, 5.2 the same as 

 those of both parents, 43.2 intermediate, 24.9 developed 

 in excess of parental extremes, and 14.1 in deficit of 

 parental extremes. It will also be seen that 17.8 per cent 

 are the same as those of one or the other parent or both 

 parents; that 82.2 per cent are intermediate and de- 

 veloped beyond parental extremes; and that 43.2 per 

 cent are intermediate against 39 per cent that are de- 

 veloped beyond parental extremes. 



Further studies of the separate percentages of macro- 

 scopic and microscopic characters show, as presented in 

 the second summary of Table I, in the former as com- 

 pared with the latter, lower percentages in the characters 

 that are the same as one or the other parent or both 

 parents and that are intermediate, but higher percentages 

 in the characters that are developed beyond parental ex- 

 tremes, especially in those which are developed in deficit 

 of parental extremes. The figures in relation to sameness 

 to one or the other parent or both parents run closely, but 

 in the other three parent-phases they show marked 

 divergence. 



The frequent absence of agreement between the dis- 

 tribution of the macroscopic and microscopic data of the 

 hybrids among the six. parent-phases is at present inex- 

 plicable. As before stated, it seems, if in any hybrid 

 given proportions of macroscopic characters would be 

 found to be the same as those of the seed parent and as 

 those of the pollen parent, that the corresponding pro- 

 portions of the microscopic characters would be found; 

 but the proportions may not only be quite different but 

 even reversed. The proportions of macroscopic and 

 microscopic characters that are the same as or inclined 

 to the seed and pollen parents, respectively, are approxi- 

 mately in Ipomwa sloteri (Table I, Summary 4) about 

 2 to 1 and 3 to 1, respectively; in Lcelia-Cattleya can- 

 hamiana, 1 to 2 and 1 to 2; in Cymbidium eburneo- 

 lowianum, 3 to 2 and nearly 1 to 1 respectively ; in Den- 

 brobium cybele, 1 to 3 and about 1 to 1 respectively; 

 in Miltonia bleuana, 4 to 3 and 1 to nearly Vfa respec- 

 tively; in Cypripedium lathamianum, about 1 to 1 and 

 nearly 1 to I 1 /*, respectively; in C. lathamianum inver- 

 sum, 2 to 1 and iy 2 to 1 respectively, and in C. nitens 

 1% to 1 and 1 to 1%, respectively. With such marked 

 and unaccountable variations of macroscopic and micro- 

 scopic values, it is to be expected that owing to the 

 great dissimilarity in the methods and characters of the 

 data of the tissue and starch investigations the two sets 

 of data may differ even more widely than the macro- 

 scopic and microscopic data just examined ; and such is 

 found to be the case, as will be shown in the following 

 section wherein additional consideration of the tissue 

 characters is given. 



3. TISSUES AND STAKCHES OF SAME 



PARENT- AND HYBRID-STOCKS. 

 COMPARISONS OF CHARACTERS OF THE TISSUES AND 

 OF THE HlSTOLOGIC AND OTHER PROPERTIES AND 

 REACTION-INTENSITIES OF THE STARCHES OF 

 HYBRID-STOCKS AS REGARDS SAMENESS, INTER- 

 MEDIATENESS, EXCESS AND DEFICIT OF DEVELOP- 

 MENT IN RELATION TO THE PARENT-STOCKS. 



(Table I, Parts 1 to 8. and Summaries 1 to 9. Charts F 1 to F 14.) 



When the present research was planned it was the 

 intention, as stated in the introduction, to make coinci- 

 dent studies of the tissues and starches of each parent 

 and hybrid specimen, with the especial object of show- 

 ing what relationships, if any, exist between the macro- 

 scopic and microscopic characters of the plants and the 

 histological and other properties and reaction-intensi- 

 ties of the starches, but various conditions combined to 

 render this project impracticable. One might be led 

 to the assumption, upon superficial thought, that if, for 

 instance, the macroscopic plant-characters of any hy- 

 brid are distributed in certain percentages among the 

 six sex-phase divisions a closely corresponding division 

 of the microscopic characters would be found, and that 

 starch characters, physical and physico-chemical, would 

 be in similar agreement. In other words, a universality 

 of type or plan of distribution of characters, so that if, 

 for example, in Ipomcea sloteri the distribution of macro- 

 scopic characters among the six parent-phases be (Table 

 I, Summary 1) 2.6, 2.6, 0, 47.4, 42.1, and 5.3 per cent, re- 

 spectively, the distribution of the microscopic characters 

 would be essentially or closely the same ; but, in fact, 

 there are more or less marked differences, as is evident 

 by the following figures for the latter: 8.4, 3.2, 2.1, 32.6, 

 47.4, and 6.3 per cent, respectively. By such compari- 

 sons it will be noted that, among the macroscopic char- 

 acters as compared with the number of microscopic char- 

 acters, less than one-third will be the same as those of 

 the seed parent (2.6 : 8.4) ; a slightly smaller percentage 

 the same as pollen parent (2.6 : 3.2) ; a smaller percent- 

 age the same as both parents (0:2.1); a very much 

 higher percentage intermediate (47.4:32.6); a smaller 

 percentage developed to excess of parental extremes 

 (42.1 : 47.4) ; and a slightly smaller percentage devel- 

 oped in deficit of parental extremes (5.3:6.5). Such 

 differences vary in the different hybrids in both quantity 

 and direction, and when the percentages for all of the 

 hybrids are summed up, as in Table I, Summary 2, the 

 macroscopic characters show distinctly higher percent- 

 ages than the microscopic characters in regard to same- 

 ness as the seed parent, pollen parent, and both parents, 

 and also to intermediateness, especially the latter; and 

 markedly lower percentages in the characters developed 

 beyond parental extremes. 



In view of such extraordinary differences in percent- 

 ages of microscopic and macroscopic characters, interest 

 is at once aroused in regard to the relative peculiarities 

 of the tissues and starches in their parental relationships. 

 On general principles it seems probable that if two 

 groups of characters which are so closely related as the 

 naked eye and microscopic characters differ so notably 

 that the group of characters consisting of reaction-inten- 

 sities of the starches should differ as much or more from 



