286 BOTANY part i 



dynamic hybrids), or they may show an equal combination of the 

 characters of both (homodynamic hybrids, e.g. Nicotiana rustica 9 

 X Nic. panimlata c? )• Less frequently it happens that the hybrid 

 resembles one ancestor almost exclusively. Thus, certain hybrids of 

 the Strawberry resemble the male parent only. In such a case the 

 attributes of the other ancestor remain latent, and may appear regularly 

 or quite unexpectedly, through atavism (p. 284), in later generations. 



The mingling of characters is often complete. Had one species simple leaves 

 and the other compound, their hybrid would have leaves more or less cleft ; or 

 \vere the flowers of one parent species red and those of the other yellow, the 

 hybrid frequently bore flowers with red and yellow markings (mosaic hybrids), or 

 which were orange-coloured. If an early blooming form were crossed with a late 

 bloomer, the hybrid would flower at a time intermediate between the two. 



A large number of spontaneous hybrids have been found which have arisen 

 naturally from plants with a special tendency to liybridisation. That such 

 natural hybrids do not occur oftener is due to the lack of an opportune time or 

 space for their development, and also to the fact that in the case of pollination 

 of flowers with diff'erent kinds of pollen, that of their own species seems as a 

 rule more eff"ectual in eff'ecting fertilisation. 



The investigations ofGnEGOR Mendel, de Vries, Correns, Tsohermak, and 

 others, have greatly increased our knowledge of the laws governing the inheritance 

 of different characters of plants. No final conclusions are yet attained, however, 

 since even in regard to limited portions of the problem new features are still 

 becoming apparent {^''). 



In the crossing of closely related forms (varieties or nearly related species), 

 Mendel's laws (called after their first discoverer, Gregor Joh. Mendel [1866]) are 

 found to hold. These laws, the most generally applicable ot which is that of the 

 segregation of characters, were independently rediscovered by de Vries, Correns, 

 and TscHERJiAK. An example will give the best idea of these laws. If a red- 

 flowered MirabiHs Jnlapa be crossed with a white-flowered individual one obtains 

 a generation of hybrids with uniformly rose-coloured flowers. If these are 

 fertilised from one another a second generation is obtained, but tlie individuals of 

 this are not uniformly coloured ; in addition to rose-coloured plants pure red- 

 flowered and white-flowered plants occur in the proportion per cent of 50 : 25 : 25, 

 i.e. in the ratio 2:1:1. When fertilised from one another the pure red-flowered 

 plants produce a red-flowered progeny and the white-flowered plants also breed 

 true ; they have returned to the pure parent forms. The 50% of rose-coloured 

 plants again splits in the next genei'ation, and like the former generation yields 

 25% pure red, 25% pure white, and 50% rose-coloured plants. The proportion of 

 hybrid plants thus continually becomes lessened by the return to the red and 

 white types ; in the eighth generation only 075% of liybrids remain, and this 

 small remainder continues to split furtlier on breeding. These results are 

 theoretically explained by assuming that the sexual cells of the rose-flowered 

 hybrids are not themselves of hybrid nature, but are already segregated into 

 pure red and pure white sexual elements. In tlie process of fertilisa- 

 O — O tion tlie union iiroducing a hybrid, red x white (white x red, red x white) 

 V will occur twice as frequently as tlie union red x red or white x white 

 ^^0 which give rise to pure forms. This is illustrated in the accompanying 

 diagram in whicli tlie dark circles represent the red sexual elements. 



The characters in wliieh the parents difler do not, liowever, always blend so 



