340 THE 'ORIGIN OF SPECIES.' [i86o. 



much when tabulating the varieties of plants, and I have a 

 discussion somewhere on this point. It is absolutely implied 

 in my ideas of classification and divergence that only one or 

 two species, of even large genera, give birth to new species ; 

 and many whole genera become wholly extinct .... Please 

 see p. 341 of the 'Origin.' But I cannot remember that I 

 have stated in the ' Origin ' the fact of only very few species 

 in each genus varying. You have put the view much better 

 in your letter. Instead of saying as I often have, that very 

 few species vary at the same time, I ought to have said, that 

 very few species of a genus ever vary so as to become modified ; 

 for this is the fundamental explanation of classification, and 

 is shown in my engraved diagram. . . . 



I quite agree with you on the strange and inexplicable fact 

 of Ornithorhynchus having been preserved, and Australian 

 Trigonia, or the Silurian Lingula. I always repeat to myself 

 that we hardly know why any one single species is rare or 

 common in the best-known countries. I have got a set of 

 notes somewhere on the inhabitants of fresh water ; and it 

 is singular how many of these are ancient, or intermediate 

 forms ; which I think is explained by the competition having 

 been less severe, and the rate of change of organic forms 

 having been slower in small confined areas, such as all the 

 fresh waters make compared with sea or land. 



I see that you do allude in the last page, as a difficulty, to 

 Marsupials not having become Placentals in Australia; but 

 this I think you have no right at all to expect ; for we ought 

 to look at Marsupials and Placentals as having descended 

 from some intermediate and lower form. The argument of 

 Rodents not having become highly developed in Australia 

 (supposing that they have long existed there) is much stronger. 

 I grieve to see you hint at the creation " of distinct successive 

 types, as well as of a certain number of distinct aboriginal 

 types." Remember, if you admit this, you give up the em- 

 bryological argument (tlie weightiest of all to me), and the 



