INTRODUCTION. 



11 



vegetable kingdom in general the male and female pro- 

 creative elements are of equal potency. The rule of the 

 similarity of reciprocal hybrids, as in all other rules in 

 the study of hybrids, is not without exceptions. It is 

 self-evident that a certain dissimilarity of reciprocal hy- 

 brids can be correctly attributed only to the stronger 

 influence of the male or of the female elements if the 

 experiments axe carefully carried out in the same way, 

 and if they have, after many repetitions, always given 

 rise to the same results. Nearly all of the reports up to 

 this time leave much to be desired in these respects and 

 room for justifiable doubt. The following statements on 

 the dissimilarity of reciprocal hybrids are worth con- 

 sideration : 



a. The female element influences most strongly all 

 parts of the morphology of Pelargonium fulgidum X 

 P. grandiflorum, P. peltatum X P- zonale, Epilobium 

 hirstdum X E. tournefortii. In many Digitalis hybrids 

 it influences most strongly the coloring of the flowers, 

 and in several the forms of the corolla also. In 

 Nymphcea rubra X N. dentata the cotyledons are always 

 much more like those of the female parent species. 



6. The female element exercises apparently a pre- 

 dominating influence on the capacity of resistance to cold 

 of Rhododendron (hybrid of R. arboreum), of Lycium, 

 and possibly also of Crinum (hybrid of C. capense). 



c. The influence of the male element is predominant 

 in all parts-of the morphology of Papaver caucasicum X 

 P. somniferum and Cypripedium barbatum X 0. villo- 

 sum (ob constant ?) . It exercises a powerful influence on 

 the flower coloration of Petunia. 



d. Gartner has several times noticed variations in the 

 fertility of the seed of the offspring in reciprocal hybrids, 

 as in Dianthus barbatus X D- superbus. Gartner's ex- 

 periments are, however, hardly suificient to prove the 

 uniformity of these flndings in the hybrids concerned. 

 (In literature there may be found many speculations 

 advanced on the influences of the male and female ele- 

 ment on the properties of a hybrid, but supported by the 

 descriptioii of only one hybrid.) It is evident there can 

 be no basis for comparison unless the forms resulting 

 from A 9 X B 5 and B 9 X A 3 are both known. 



Departures of an isolated specimen of a hybrid from 

 the typical form are much more frequently noticed and 

 are entirely independent of the rSles played by the parent 

 forms in their production. Not infrequently, important 

 differences appear in seedlings from a single crossing that 

 are grown under absolutely similar conditions. These 

 variations show themselves in various ways. 



a. Individuals resulting from a given hybridization 

 show among themselves unimportant differences, espe- 

 cially in the coloring of the flowers and other similarly 

 easily altered characteristics, as in the hybrids of Ver- 

 bascum phmniceum, Salix cuprea X S. daphnoides. 



b. The hybrid appears in two different types, each 

 showing a different combination of the characters of the 

 parent species. As a rule, the one type is closer to one, 

 and the other to the other, parent species ; the frequency 

 of the appearance of both types is often very variable. 

 Gartner designated the type which appears less fre- 

 quently as the exceptional type (" Ausnahmetypus"). 

 Instances may be seen among Cistus, Dianthus, Gmm, 



Oenothera, Lobelia, Verbascum thapsus X I^. nigrum, 

 Nicotiana quadrivalvis X N- tabacum macrophylla. 



c. The hybrid appears in several different types. 

 Gartner gives several examples of this, but there are 

 only three known forms by a polymorphic union. 



d. The hybrid shows one typical form of a mid- 

 intermediateness, together with a number of varying 

 forms that are usually closer to one or the other parent, 

 among which no well-marked types can be distinguished. 

 Such is the behavior of Medicago falcata X M. sativa, 

 and similarly of Melandryum album X M. rubriim. 



e. The hybrid is polymorphous from the beginning. 

 The observations up to the present leave it doubtful 

 whether one should in these circumstances distinguish 

 between varying forms or between several fixed types 

 with similar combinations of properties. Examples: 

 Abutilon, hybrids of Pelargonium glaucum L'Her., P. 

 radvJa X P- myrrhifolium, Passiflora, Hieracium, Ne- 

 penthes, Narcissus. Gartner has offered the hypothesis 

 that hybrids between different species are always of the 

 same form and that the hybrids between varieties are 

 polymorphic. If by " varieties " garden forms or garden 

 hybrids are understood, this rule is correct; but if, on the 

 other hand, one understands constant races of pure de- 

 scent it is decidedly incorrect. 



Comparisons of hybrids which arise from the same 

 species, but which are produced and grown in different 

 places, exhibit many other results. Spontaneous or natu- 

 ral hybrids are, as a rule, more variable than those pro- 

 duced artificially, as for example, Verbascum lychnitis X 

 V. ihapsus and Y. lychnitis X V. nigrum. My own hy- 

 brids between Digitalis purpurea and D. lutea were very- 

 much like one another when I sowed the seed, but a great 

 variety of forms appeared if the seeds had by chance 

 sown themselves. It may be that in these cases there is 

 no real causal connection between the varieties of the 

 forms and the methods of sowing; but, on the other hand, 

 it is a fact that different cultivators in crossing the same 

 species have very often obtained different products. 

 Hence, while similarity of the forms of all the plants of 

 one crossing appears to be without doubt the rule in 

 experiments in cultivation, similarity appears to be the 

 exception in nature. It remains to be determined how 

 great an influence dissimilar nutrition of the parent- 

 species or of the hybrid embryos may have on the varia- 

 bility of form of the hybrids. 



Second Pboposition. 



The properties of the hyhrids are derived from the properties of 

 the parents. For the most part the hybrids differ from 

 their parents .only in size and luayariance of growth and 

 in their generative poioers. 



The methods and modes in which the properties of 

 the parent species are combined in the hybrids are very 

 variable. In general, a blending or mutual penetration 

 of the different properties is found, often in such a way 

 that in one respect the one and in another the other 

 parent form appears to predominate. That is to say, 

 in many instances the hybrid resembles one parent more 

 in the leaves, and the other parent more in the flowers. 

 Now and then an exceptional variety of the hybrid (the 

 "Ausnahmetypus" of Gartner) appears in which the 

 properties are inversely apportioned. Many hybrids at 

 first more nearly resemble one, and later more nearly 



