266 



2. iTi-eat physical and climatic differences, even in comparatively narrow 

 limits. No more striking illustration can be imagined than is offered by the 

 streams of the Pacific slope of North America, which are inhabited by extraord- 

 inarv variable species, without stable varieties 



3. Amphimyxis has been suggested by Ayres as a condition favoring the 

 display of great variation. 



These are simply statements under whicii variation seems to find its optimum 

 condition and do not approach any explanation of its causes. 



CLA>iSlFiCATiON? OF VARIATIONS. — Students of variation have found it ad- 

 vantageous to analyze the phenomena, and the result of this analysis has given us 

 the following classifications: 



Continuous variation, including all gradual modifications and transitions. 



Discontinuous variation ; any sudden and wide modifications or saltations. 



Using other features as the basis of classification, we have : 



Meristic variations dealing with the change in the number of successive parts. 



Substuntative dealing witli the chemical modifications of parts. 



Another classification gives us: 



Indeterminate, or fortuitous and aimless variation. This is largely individual 

 and pertains to series of variations either geographically or geologically. 



iJi'tcrminate and adaptive, leading to definite end. 



The most essential and at the same time the most difficult to define is the 

 distinction between — 



Ontogenetic variation including all those deviations appearing at any time, 

 from any cause, during the life cycle of an individual; 



Plii/lo(jenetic variations change from the specific characters appearing at some 

 time in the life cycle of an individual, or better still, a large number of indi- 

 viduals, reappearing in the next generation, finally becoming hereditarily fixed. 



I have in the following directions omitted the use of the terms ontogenetic 

 and phylogenetic. Recently (Osborn, 1894 , the distinction between ontogenetic 

 and phylogenetic variation in the study of evolution has been strenuously insisted 

 upon as the only possible way of determining the value of any given variation in 

 the process of ev<dution. However, it is certainly intpossible in many cases to 

 determine whether a given v:iriation is ontogenetic or phylogenetic as defined by 

 Osborn. To give a concrete case. The ancon sheep of evolutionary classics was 

 born with short legs. Were they ontogenetic or phylogenetic? Subsequent events 

 proved that they were phylogenetic, but certainly the short legs in themselves 

 enabled no one to make the distinction; the hereditary transmission decided the 



