senting them. If there be several njaxima suited for 

 nearly the same province, the maximum of greatest intensity 

 will choke out the others. So, too, there are probably 

 many maxima now unoccupied, as for instance, the thistle 

 represented a m^aximum of vegetable life in South America, 

 but till man imported the thistle to fill it up, other maxima 

 of less intensity held the ground. In some cases possibly 

 several maxima are closely related, and differ little in their 

 intensity, so that slightly differing species exist together, 

 and may in their variation pass one into the other, as 

 perhaps in brambles and some species of St. John's wort, &c. 



If then the province of a species, i. e. the physical geo- 

 graphy of a country alter, and its enemies and food with 

 them, clearly the maximum will shift and the species change. 

 But this is not the evolution of new species, though to a 

 person who only notes geological evidence it appears so. For 

 just as in a storm the lightning shews the trees still, though 

 really waving to and fro, so the different species in geology 

 are probably but steps in a constant change. Such a change 

 of course must be slow for life to follow it, for a species con- 

 sists quite as much in a bundle of acquired and transmitted 

 habits as in a certain formation of organs, and the change in 

 habit will probably be far slower than the change in form. 



How then do new species arise ? For we see that, if the 

 species be a maximum of vitality, in a multitudinous progeny 

 those nearest the type will choke out the others and the 

 species will be stable. Varieties will be connected with 

 maxima of vitality in two ways. Firstly, slight differences 

 in the province will slightly shift the maximum. Thus 

 mountain sheej) v/ould be more agile than low land sheep. 

 Secondly, in such a way as tliis. Suppose this table 

 a low mound, narrow though long. Then the height at any 

 point will be a function of the distances from the N. and E. 

 walls of the room. There will be one point of maximum 

 height, but whilst a change N. or S. produces a great change 

 in the altitude, one E. or W. will produce but little. So 



