322 TAXONOMY. 



within the species, and especially in the application of the pollen 

 to the stigma of the same blossom. Commonly the sterilit}* of 

 hybrids is owing to the impotence of the stamens, which perfect 

 no pollen ; and most such hybrids may be fertilized by the pollen 

 of the one or the other parent. Then the offspring either in 

 the first or second generation reverts to the fertilizing species. 

 Moreover, certain hybrids, such as those of Datura, which are 

 fully fertile per se, divide in the offspring, partly in the first gen- 

 eration, and completely in two or three succeeding generations, 

 into the two component species, even when close-fertilized. 1 (In 

 part this may come from adventive embryo-formation, 533.) 



639. There appears, therefore, to be a real ground in nature 

 for species, notwithstanding the difficulty and even impossibility 

 in many cases of defining and limiting them. 



640. Species is taken as the unit in zoological and botanical 

 classification. Important as varieties are in some respects, 

 especially under domestication and cultivation, they figure in 

 scientific arrangement only as fractions of species. Species 

 are the true subjects of classification. The aim of sj^stematic 

 natural history is to express their relationship to each other. 



641. The whole ground in nature for the classification of spe- 

 cks is the obvious fact that species resemble or differ from each 

 other unequally and in extremely various degrees. If this were 

 not so, if related species differed one from another by a constant 

 quantity, so that, when arranged according to their resemblances, 

 the first differed from the second about as much as the second 

 from the third, and the third from the fourth, and so on, or if 

 the species blended as do the colors of the rainbow, then, with 

 all the diversitj" in the vegetable kingdom there actually is, there 

 could be no natural foundation for their classification. The mul- 

 titude of species would render it necessary to classify them, but 

 the classification would be wholly artificial and arbitrary. The 

 actual constitution of the vegetable kingdom, however, as ap- 

 pears from observation, is that some species resemble each other 

 veiy closely indeed, others differ as widely as possible, and be- 

 tween these the most numerous and the most various grades of 



1 According to Naudin in Comptcs Rendus, xlix. 1859, & Iv. 1862. See 

 also Naudin's memoir on hybridity in plants in Ann. Sci. Nat. ser. 4, xix. 

 1863, pp. 180-203, & in Mem. A cad. S^-i. . . . For the literature on vegetable 

 hybrids, see Kcelreuter, Nachricht, &c., 1761, and Appendices, 1763-1766; 

 Herbert, on Amaryllidaceae, 1837 ; C. F. Gasrtner, Versuche und Beobachtun- 

 gen ueber die Bastarderzeugung in Pflanzenreich, 1849; Wichura, Die 

 Bastardbefruchtung im Pflanzenreich, erlautertert an den Bastarden der 

 Weiden ; and the mnmoir of Naudin referred to. 



