INTRODUCTION. 



13 



two species whose flowers are of different size, those of 

 the hybrid are frequently of the same size or approxi- 

 mate the size of the bloom of the species having the 

 larger flowers. Examples of uncommonly large flowers 

 are seen in Dianthus arenarius X D. superbus, Rubus 

 ccesius X R- bellardii, hybrids of Rosa gallica, Begonia 

 boliviensis and Isoloma tydceum. 



A high vegetative power is very common in hybrids, 

 as in Nymphwa, Rubus ccesius, Nicotiana suaveolens X 

 N. latissima, Linaria striata X L. vulgaris and Potamo- 

 geton. A greater duration of life has been noted in con- 

 nection with several hybrids of Nicotiana and Digitalis. 

 An increased resistance to cold has been noted especially 

 in Nicotiana suaveolens X N. tabacum latiss.; while, on 

 the other hand, Salix viminalis X S. purpurea is more 

 sensitive to cold than either parent species. 



These facts point in part to an apparent lessened 

 vitality of hybrids in consequence of their abnormal mode 

 of production ; and in part in some instances to an extra- 

 ordinary vegetative power. The cause of this last phe- 

 nomenon, which is observed less frequently than lessened 

 vitality, has been in some degree only recently under- 

 stood. Noteworthy experiments of Knight, Lecoq, and 

 others have been published, but it has been through the 

 painstaking researches of Charles Darwin that the ease 

 with which a cross between different individuals and 

 races of one and the same species is effected was first 

 clearly explained. The increase of the vegetative power 

 in hybrids is clearly a phenomenon that closely corre- 

 sponds with the peculiar conditions of hybrid produc- 

 tion, and needs not a special explanation. It was at first 

 thought that lessened fertility was compensated for by 

 greater vegetative luxuriance, an hypothesis that Gart- 

 ner has shown to be untenable, as is evident by the fact 

 that many of the most fertile hybrids (Durata, Mirabilis) 

 are also notable for the largest growth. 



4. PARTIAL OR COMPLETE STERILITY OF HYBRIDS. 



Subnormal fertility of hybrids, especially as regards 

 the pollen, has long been recognized as one of the most 

 important criteria of hybrids. It seems, however, that 

 this character like intermediateness has been an almost 

 unbridled conception and hence greatly overvalued as a 

 distinguishing feature. Focke in his summary gives us 

 a wealth of facts in this connection : 



FOURTH PROPOSITION. 



Hybrids between different species show in their anthers a 

 smaller number of normal pollen-grains and a, smaller 

 number of normal seed than in plants of pure descent. 

 Frequently they produce neither pollen nor seed. In 

 hybridization between nearly related races this weakening 

 of the power of sexual reproduction is not present. The* 

 flowers of sterile or nearly sterile hybrids usually remain 

 fresh for a. long time. 



No property of hybrids has attracted so much atten- 

 tion as the lessening of the ability of sexual reproduc- 

 tion. Kb'lreuter believes that this peculiarity permits 

 a sharp border-line to be drawn between species and 

 varieties. Since then many botanists have accepted the 

 same view, and lately B. Naudin, Decaisne, and Caspary 

 have adopted it in a more or less modified form. Knight 

 and Klotzsch, and before them Godron, hold that the 

 pollen of hybrids is entirely impotent, which contention 



had already been disproved by Kolreuter's accurate re- 

 searches. Kolreuter is accredited with the promulga- 

 tion of the doctrine of complete sterility of hybrids, but 

 this erroneous charge is to be explained only through 

 an ignorance or misunderstanding of the Latin texts: 

 Kolreuter does not speak of complete sterility, but only 

 of a lessened fertility, as a universal property of hybrids. 



In different plant genera the fertility of hybrids is 

 very varied. Fertility is observed in a very low degree 

 in the hybrids Papaver, Viola, Verbascum, and Digitalis; 

 it is more common in Anemone, Nicotiana, Mentha, 

 Crinum, Cucurbitacea, and Passifloracece ; and it is more 

 common than sterility in Aquilegia, Dianthus, Pelargo- 

 nium, Gcum, EpUobinm, Fuschia, Cotyledon, Begonia, 

 Cirsium, Erica, Rhododendron, Calceolaria, Quercus, 

 Salix, Gladiolus, Cypripedium, and Hippeastrum. In 

 the genera Vitis, Primus, Fragaria, and Pirus, hybrids of 

 closely related species are used as seed-bearing plants; 

 and in Cereus the hybrids of widely separated species 

 show undiminished fertility. 



The sterility of hybrids is expressed at times by their 

 showing no inclination to flower, which peculiarity has 

 been noticed especially in several hybrids of Rhododen- 

 dron, Epilobium, Cereus, and Hymenocallis; but these 

 are exceptions, inasmuch as hybrids usually flower more 

 abundantly and earlier than true species. 



In hybrids with unisexual flowers the male flowers 

 fall off when in the bud, as in Cucurbitaceo? and Be- 

 gonia (hybrids of B. fmbeli A. DC.). In bisexual flowers 

 the stamens are stunted, as noted in several hybrids of 

 Pelargonium and Digitalis (D. luteaXD. purpurea f. 

 fubiflora Lindl.). The most common sequel of hybrid 

 production is a deficient development of the pollen-grains 

 in hybrid plants. Commonly the anthers of hybrids are 

 sterile and do not contain any pollen; or they are 

 small and do not open. Such deficiency of pollen is 

 noted in Rubus idcrm X #. odoratus, Ribes aureum X 

 R. sanguineum, and Alopccurus geniculatus X A. pra- 

 tensis. In other cases the stamens produce small pow- 

 dery grains which do not swell with moisture, which are 

 of varying size and shape, and with which are usually 

 mixed a few single, well-formed, embryo-forming pollen 

 grains. The number of normal grains is, however, fre- 

 quently larger, and comprises 10, 30, or more per cent 

 of the total number. Large, rough grains which swell 

 with moisture, together with small well-formed grains, 

 are present often in greater or less number among the 

 stunted grains. In hybrids of closely related species, as 

 in Melandryum album X M. rubrum, but little irregu- 

 larity is usually found in the form of the pollen-grains. 

 In one hybrid, Sinningia, the pollen was better in the 

 second year of flowering than in the first. 



In the hybrids of unquestionably different species a 

 normal formation of the stamens is seldom met with. 

 Assertions in support of this still need confirmation, in 

 part, therefore I refer to Nymphaa lotus X N. rubra, 

 Begonia rubrovenia X B. xanthina, Isoloma tyda:um X 

 1. sciadocalyx X Salix purpurea X S. repens; pollen 

 grains which are all of nearly the same form are found 

 in Salix aurita, and S. caprea and S. viminalis X S. 

 repens. 



On the other hand, a deficient development of stamens 

 appears less frequently in race crossings. Possibly, fur- 



