November ii, 1915] 



NATURE 



285 



jjix'ssly not stated in my address, as quoted by my 

 reviewer, "that there is no evidence whatever in sup- 

 port of the theory that ' species ' may arise from 

 fortuitous, saltatory characters." No such sweeping 

 statement is made ; on the contrary, the much more 

 i^uarded statement is made (p. 229) that, " I do not 

 know of a single instance where a field observer in 

 mammalogy or in palaeontology has recorded a new 

 >altation [i.e. mutation] character which is known to 

 !)(' of any significance in the evolution of the race. . . . 

 Saltation is, however, theoretically probable in certain 

 numerical and meristic characters, such as super- 

 numerary teeth and vertebrae." 



Nor has the reviewer understood my observations 

 regarding the nature of species and genera. The 

 Darwinian tradition as to the actual modes of evolu- 

 tion of species, which is closely followed by De Vries 

 and Bateson to-day, may be described as a " single 

 character" or " group character" doctrine, namely, 

 that the evolution of species manifests itself suddenly 

 in one character or group of characters; that when- 

 ever such character or characters arise through in- 

 dividual variation or through saltation, they may be 

 fixed in the race by selection. This was one of the 

 essential features of our former conception of evolu- 

 tion, namely, that an organism advances now here, 

 now there ; it was drawn from ancient traditions about 

 the nature of "species" entirely prior to the period of 

 palaeontological discovery of the' actual modes of evolu- 

 tion of animal forms. These date back to the discovery 

 of mutation by the German palaeontologist Waagen in 

 i86g, mutation in an entirely different sense from that 

 of De Vries. Palaeontology has now demonstrated 

 that we must abandon this "single character" con- 

 ception of evolution, because it proves that every 

 organism is made up of an almost infinite number o'f 

 characters, each of which is in a continuous state of 

 movement, either originative, progressive, or retro- 

 gressive. This is a law well observed and established 

 through palaeontology, for which at present we offer 

 no theoretical explanation whatever. The "single 

 characters" and "group characters" of the older 

 systematists and of Darwin are now seen to be those 

 which, although part of a large general advance or 

 retrogression, happen to be so conspicuous as to attract 

 the eye of the zoologist. In my address this newlv- 

 discovered law of multiple-character evolution is show'n 

 to be entirely in accord with the theory of the "unit- 

 character" constitution of the germ, which chiefly 

 dates from Mendel's discoveries. 



In regard to natural selection, I am not convinced, 

 as my reviewer states, "that all characters must run 

 the gauntlet of selection." On the other hand, I give 

 it as my theoretic opinion (p. 239) that "Selection is 

 operating always upon the sum of all the movements, 

 actions, and reactions of characters known as the 

 organism and upon all single characters [which happen 

 to be] of survival or elimination value." 



Finally, I cannot lend the least shadow of assent, 

 as my reviewer states I do, to the recent enunciation 

 by Prof. Bateson that we may have to forgo the 

 theory of addition of germinal determiners and sub- 

 stitute a theory of evolution bv loss of determiners. 

 On the contrary, I express^ state that while 

 palaeontology offers the most positive testimonv as 

 to evolution by the loss of certain somatic characters, 

 such as in the side digits of the horse (of which, by 

 the way, most of the germinal determiners still per- 

 sist), the law of incessant additions of new numerical 

 and proportional characters such as transform the 

 Eocene into the modern mammals certainly does not 

 accord with Bateson 's metaphysical conception that 

 this outward progress and development may represent 

 an inward loss of germinal determiners. The defect 



NO. 2402, VOL. 96] 



in Bateson 's reasoning arises from the fact that in 

 his special field of investigation the new laws and 

 principles established in palaeontology are not observ- 

 able at all. 



Of the three kinds of mammals mentioned above 

 the group at present revealing the most complete and 

 unbrdken history is the titanothere family, which I 

 have been engaged in monographing during the past 

 fifteen years for the United States Geological Survey. 

 The most significant result derived from the intensive 

 study of the evolution of this family is the law of 

 multiple-character evolution, namely, that evolution 

 simultaneously progresses in every one of the in- 

 numerable single characters of which an organism is 

 composed. Each character possesses its own indivi- 

 duality and separableness, each advances or lags 

 behind according to its individual velocity, each sub- 

 serves the general purpose of the entire organism 

 through the uniting forces which we term correlation. 

 While writing, I would like also to correct an 

 erroneous impression, for which my present reviewer 

 is not at all responsible, as to my viewpoint regarding 

 Lamarckism. In 1889 I held that the reaction- 

 inheritance hypothesis of Lamarck was the only ex- 

 planation we have for such directive or orthogenetic 

 tendencies in evolution as are displayed especially in 

 the origin of characters which I have termed "recti- 

 gradations." I subsequently — with Baldwin and Lloyd 

 Morgan — discovered the organic-selection principle 

 and abandoned the reaction-inheritance interpretation 

 when I observed that while it may explain certain 

 orthogenetic tendencies, it is directly contrary to 

 others. Thus I am to-day in no sense a Lamarckian. 

 All that we can positively state at present is that 

 the adaptive trend in the origin of new specific and 

 generic characters is the resultant of the interactions 

 of four sets of forces : germinal, environmental, 

 ontogenetic, and selective, as expressed in the law of 

 " tetraplasy." We know that these "actions," to quote 

 Newton's phrase, are in continuous operation. We 

 also know that in certain cases one or other of 

 these four complexes of forces may be dominant, so 

 that the visible changes which we observe in animals 

 and plants may for a time be chiefly either ontogenic, 

 or environmental, or germinal, or selective. But what 

 may be the nature of the interactions between onto- 

 genic, environmental, and selective forces on the germ- 

 chromatin and the continuous origin of new adaptive 

 characters in the germ-chromatin we have not as yet 

 the most remote conception. 



Finally, the chief purpose of my address was to 

 show the real harmony which exists between the 

 actual observations of palaeontologists and of experi- 

 mentalists and the distinctive principles which are 

 most readily observed in these respective fields of 

 research. Henry Fairfield Osborn. 



Columbia University, September 29. 



Prof. Osborn would have earned the gratitude of 

 us all if he had produced some evidence in support of 

 his assertion that "there is a clear genetic distinction 

 between ' numerical ' and ' proportional ' characters." 

 We still maintain that this connection cannot be 

 demonstrated. 



Neither can the reviewer admit that he has failed 

 to understand Prof. Osborn 's observations regarding 

 the nature of species and genera. In the first place, 

 Prof. Osborn 's interpretation of the "Darwinian 

 tradition," which he emphasises in this protest, is 

 Inadmissible, since it involves the assumption that the 

 evolution of species manifests itself suddenly " in one 

 character or group of characters." From his context 

 he implies that according to this " tradition " a species, 

 A, may give rise to species B, C, and D, each differ- 



