14 



INTRODUCTION. 



ther research will show that it actually appears more 

 often. The only two examples that I know are in my 

 Anagallis cross-breeds. It is doubtful whether Raphanus 

 sativiis and B. raphanistrum should be considered as 

 representing species or races. It seems, however, that 

 some individual hybrids of closely related species are 

 entirely sterile, as in Capsella rubella X C. bursa pas- 

 toris, Viola alia X V. scotophylla, Papaver dubium X 

 P. rhoeas. 



Fertility of the female organs is not, as a rule, so 

 much weakened in hybrids as is that of the male organs. 

 It is, however, usually impaired to a great degree. Many 

 hybrids never develop fruit. Assertions as to the absolute 

 sterility of hybrids can not, however, be advanced without 

 manifold researches. From the crossing Rubus ccesius 

 X R. idceus one sees many thousand flowers remain ster- 

 ile and only here and there individuals produce fruit. 

 See also Digitalis lutea X D. purpurea, Lobelia fulgens 

 X L. sypliilitica, Crinum capenSe X C. scabrum. A 

 morphologically recognizable imperfection of the ovule 

 has heretofore rarely been seen, unless by Bornet in 

 Cistus. To obtain conclusive information as to .the 

 female fertility of a hybrid, the stigma should be fer- 

 tilized with pollen from the parent species, which fertili- 

 zation imiversally brings forth better fruit than the pollen 

 of the hybrid which is weakened in its fertilizing power. 

 In some cases hybrids having the pollen which has a 

 subnormal potency produce normal fruit with parental 

 pollen, as in Luffa. 



Several hybrids drop their unwithered flowers with 

 fully formed calyx and stamens, as in Ribes, Nicotiana 

 rustica X N. paniculata and other hybrid Nicotianas. 



As a rule, the corolla withers in a normal manner 

 after a longer existence than in the parent species, or it 

 will be thrown off as in the parent species ; but following 

 this there is no setting of fruit or a setting of only poor 

 fruit. In many cases the fruit while externally well 

 formed is seedless. In many other cases the fruit is set, 

 but in smaller number and with fewer seeds than in the 

 parent species. In hybrids of very closely related species 

 the number of seeds appears to be somewhat less than 

 in the parents. Examples of this, according to Gartner, 

 are Melandryum album X M. rubrum, and Lobelia car- 

 dinalis X L. fulgens. It is also true in race-crossings 

 of Verbascum. 



Hybrids of essentially different species seldom show 

 an undiminished fertility. However, no striking les- 

 sening of fertility has been observed in Brassica napus X 

 B. oleracea, Dianthus cliinensis X D. plumarius sibiricus, 

 Pelargonium pinnatum X P- hirsutum,, Abutilon, Medi- 

 cago, several Cereus and Begonias, Hicracium auranti- 

 cum X H. echioides, Nicotiana alata X N. larigsdorffii, 

 several hybrids of Erica, Calceolaria, Isoloma, Veronica, 

 and several Orchidacese. Also, among many wild-grow- 

 ing hybrids one finds fruits and seeds in great quantities, 

 as in many Rosa, Epilobias, Fuchsias, Cirsiei, Hieraciei, 

 Salices, Lobelia, and so forth. In such cases, therefore, 

 it is not sufficient to ascertain whether the plants in ques- 

 tion are primary hybrids or whether, as is usually the 

 case, they belong to later generations or have arisen 

 from back-crossings. 



In order to produce seeds or to obtain a luxuriant 

 progeny some hybrid plants require fertilization with the 



pollen of others, as in hybrids of Cistus, Begonia, Gladi- 

 olus, and Hippeastrum. 



In many hybrid plants only the first flowers produce 

 seeds, as in Aquilegia, Dianthus, Silene, Lavateria Thur- 

 ingiaca X T. pseudolbia, and Rubus foliosus X R- 

 sprengelii. In other cases the first flowerings are usually 

 sterile while the later flowerings are frequently fertile, 

 as in Datura, Nicotiana rustica X N. paniculata, N. rus- 

 tica X N. quadrivalvis, and Mirabilis. In long-lived 

 plants, the flowers in general are sterile during the first 

 year, while later, when the plant has reached a definite age, 

 they produce fruit. This is noted in Rubus idceus X R- 

 ccesius, R. bellardii X R- ccesius, Calceolaria integrifolia 

 X C. plantaginea, and Crinum capcnse X C. scabrum. 



The fertility of the ovule is, as a rule, diminished 

 to a somewhat less extent than the fertility of the pollen, 

 but there are some known examples of an opposite char- 

 acter, as in Nymphcea lotus X N. rubra, Ciconium X 

 Dibrachya in the genus Pelargonium, Lobelia fulgens 

 X L. sypliilitica, Verbascum thapsiforme X V- nig rum, 

 Narcissus montanus, and so forth. These 'are certainly 

 only of an occasional occurrence. 



The long persistence of the blossoms (especially those 

 with stamens) in many sterile hybrids corresponds with 

 the longer duration of unfertilized or incompletely fer- 

 tilized flowers. Frequently the fruit of sterile hybrids, 

 especially after fertilization with the pollen of the 

 parents, develops more or less strongly without producing 

 any seed, or producing only imperfect seeds. Especially 

 well-developed but seedless fruits are found in the Cac- 

 tacese, Passifolacese, Cucurbitacese, and Orchidaceae. 

 Gartner has studied carefully these phenomena, but in 

 the study of hybrids they hardly possess a great value. 

 Apart from this they furnish an important demonstration 

 of the correctness of the principle that the normal de- 

 velopment of the pericarp follows upon the stimulation 

 when the germinating pollen is discharged on the stigma, 

 but which is, nevertheless, entirely independent of the 

 ripening of the egg cells and the development of the 

 embryo and the seeds. 



The rule in general is that hybrids of closely related 

 races are on an average more fertile than those of defi- 

 nitely separated species. The rule can also be stated, as 

 shown above, that closely related species can more easily 

 produce hybrids than widely separated species. Both 

 rules, however, have only conditional values, for if it 

 should be concluded from this that the more easily hy- 

 brids are produced the more fertile they are, one would 

 fall into error. There is no known or traceable connec- 

 tion between the ease of production and fertility of the 

 hybrids. 



From the teleological standpoint the sterility of hy- 

 brids was formerly considered the means whereby species 

 were kept separate. Just what advantage such separa- 

 tion is (unless it be for the conveniences of the systemat- 

 ists) was never demonstrated. On the other hand, it 

 may now be asked whether or not the genesis and differ- 

 entiation of species are not brought about by the lessened 

 fertility of mongrels between well-marked races of the 

 parent type. The notable similarity between illegiti- 

 mates and hybrid offspring do not offer a basis for fur- 

 ther investigations of the causes of sterility. A better 

 explanation is probably afforded by the hybrids of Equi- 



