748 



SCIENCE 



[N. S. Vol. XXVI. No. 674 



proved by vise, but, on the contrary, are de- 

 stroyed by use. 



We note (4) that there appears to be an 

 analogy between heredity and ontogeny. 

 This hurrying up or acceleration of char- 

 acters in heredity parallels the acceleration 

 of useful characters in ontogeny. In other 

 words, from unknown causes (even if the 

 Lamarckian inheritance is admitted) char- 

 acters are accelerated (hastened) or re- 

 tarded (slowed up) in development, accord- 

 ing to the needs of the animal. Thus there 

 arises this most interesting analogy be- 

 tween the hereditary origin of new char- 

 acters and the subsequent ontogenetic his- 

 tory of characters after they have reached 

 a presumably adaptive condition. In 

 other words, just as the lateral digits of the 

 horse are retarded and the median digits 

 are accelerated, so the origin from un- 

 known causes of new characters is acceler- 

 ated or retarded, according to the needs of 

 the animal. For example, a postero-inter- 

 nal cusp of the upper grinding teeth, 

 known as the hypocone, and the interme- 

 diate cusps, known as the conules, are 

 retarded in hereditary origin in insectivores 

 and in frugivorous animals; they are ac- 

 celerated in origin in herbivorous animals. 

 We note (5) that our failure to see any 

 reasons or causes of these timely hereditary 

 origins of new characters has no bearing 

 whatever on the fact of the existence of 

 such origins, that fact is a matter of direct 

 observation independent of hypothesis. 

 For my own part I have for many years 

 (ever since I observed this fact and recog- 

 nized all- the difficulties in the Lamarckian 

 explanation) stood as a complete agnostic 

 as to the cause of such origins. I now re- 

 peat that we have no conceivable explana- 

 tion at present. 



We note (6) that an important distinc- 

 tion must now be made, namely, that such 

 origins of new characters are chiefly nu- 

 merical; something is added to the organ- 



ism which did not exist before, the rudi- 

 ment of a cusp, or the rudiment of a horn. 

 The changes of form of proportion and of 

 modeling, follow after. 



A very interesting thought has just come 

 to me during the preparation of this paper, 

 a paper which summarizes the conclusions 

 I have been gradually forming in the last 

 twenty-one years. The thought is this: 

 That theoretically there is no conflict be- 

 tween the hypotheses of continuity and dis- 

 continuity. If there does exist hereditary 

 predisposition to evolve in a definite direc- 

 tion, it may manifest itself suddenly^ as a 

 saltation, or a "mutation of de Vries," or 

 very gradually as a reetigradation, or 

 "mutation of Waagen." 



SUMMARY 



The following facts are those which are 

 put forth through paleontological observa- 

 tion, for verification by others : 



1. That many origins of new characters 

 are through some internal action in 

 heredity. 



2. That many important adaptive char- 

 acters arise determinately, definitely, but 

 by extremely slow stages. 



3. That degrees of similarity in such 

 origins correspond with degrees of kinship. 



4. That degrees of kinship also affect to 

 a certain extent, but not absolutely the 

 time of appearance, or the time of the 

 origin, or the rate of evolution. 



5. That such origins find expression not 

 spontaneously, or irrespective of condi- 

 tions, or from purely internal mechanical 

 causes, but through some entirely unknown 

 and at present inconceivable relation to 

 ontogeny (habit and use), to environment 

 (external conditions), and to selection. 



6. That if such origins do spring from 

 internal hereditary principles, as they ap- 

 pear to do in many cases, slow origins 

 (mutations of de Vries) may be simply 



