When a young naturalist begins studying an unfamiliar group of organisms, he will find it difficult to determine differences that delineate species and those that delineate varieties, because he doesn’t yet know the amount and kind of variation within the group. (This demonstrates, at least, that there is often some variation.) But if he confines his attention to one class in one region, he will soon make up his mind about how to classify doubtful forms. He will define many species because the differences impress him, and he lacks a general knowledge of analogical variation in other groups and other countries. As he extends his range of observation, he will discover more cases that are difficult to classify, for he will encounter more forms that are closely related. If his observations are extended further, he will ultimately make up his mind about species and varieties, but only at the expense of admitting extensive variation – an admission other naturalists will dispute. When he comes to study related forms from regions that are no longer continuous, and where intermediate links between doubtful forms will necessarily be absent, he will have to trust almost entirely to analogy and his difficulties will rise to a climax.
No clear line has been drawn between species and sub-species (forms that are nearly ranked as species but not quite there), sub-species and varieties, or lesser varieties and individual differences. They all blend into each other in a smooth series, and a series impresses the mind with the idea of an actual progression.
Although individual differences are unimportant to the systematists, they are very important to this argument as the first step toward the minor varieties barely thought worthy of notice in annals of natural history. I consider even slightly distinct and permanent varieties as steps leading to still more distinct and more permanent varieties, and these in turn as leading to sub-species and species. Progression from one stage to another may in some cases be due merely to the continuous and long-term action of different environmental conditions in two different regions, but I have little faith in this view. Instead, when a variety changes from differing little from its parent to one that differs more, I attribute the change to natural selection having accumulated structural differences in certain definite directions. Therefore, I believe a distinct variety can be called an incipient species. Whether this is justified should be judged by the general weight of observations and ideas presented throughout this book.
Not all incipient species actually become species. They may become extinct in the incipient state or simply remain varieties for very long periods (as Mr. Wollaston has shown for certain mollusk varieties based on fossil land shells in Madeira). If a variety were to flourish and exceed its parent species in number, it would be ranked as a species and the parent as a variety; it might supplant and exterminate the parent or coexist with it, both being ranked as independent species. (I will return to this subject.)
From these considerations I take the term “species” as being arbitrarily assigned for the sake of convenience to groups of individuals closely resembling one another. It does not essentially differ from the term “variety,” given to less distinct and more fluctuating forms. “Variety” is also a term that is applied arbitrarily and for convenience, again, in comparison with mere individual differences.
Guided by theoretical considerations, I thought that tabulating all the varieties of several well-studied plants would yield some interesting data about the nature of species that vary the most. At first this seemed simple, but Mr. H. C. Watson and Dr. Hooker soon convinced me of the many difficulties. (I am reserving the tables themselves and a discussion of these difficulties for my later work.) After carefully reading my manuscript and examining the tables, Dr. Hooker agreed that the following statements are fairly well established. Nevertheless, the subject is complex and treated tersely here, so allusions cannot be avoided to “the struggle for existence,” “divergence of character,” and other concepts that will be discussed later.
Alph. de Candolle and others have shown that plants with extensive ranges generally have varieties. This might have been expected because it means exposure to diverse physical conditions and competition with different sets of organisms (competition is by far the more important, as discussed later). My tables also show that in a limited region the most common species (those with the greatest number of individuals) and the species that are most diffused often have varieties well-defined enough to be recorded in botanical works.2 Therefore the most “dominant” species – those that range most extensively over the world, are the most diffused in their native regions, and are the most numerous – are the ones that most often produce varieties, or as I consider them, incipient species. This too might have been anticipated, because for a variety to become permanent, it must struggle with other inhabitants of the region. Already dominant species are the most likely to leave offspring, which, in addition to being slightly modified, will inherit those advantages that enabled their parents to become dominant.
If the plants of a particular region are separated into two groups, with those belonging to large genera in one and those belonging to small genera in the other, then a greater number of dominant species will be found on the side of larger genera. Again, this might have been anticipated: if many species of a genus inhabit a region, it shows that something about the organic or inorganic conditions of that region are favorable to that genus; therefore a large genus encompassing many species contains a proportionally greater number of dominant species. But so many factors tend to obscure this result that I am surprised my tables show even a small majority of dominant species among large genera. To give examples of how complex this is, consider that freshwater and salt-loving plants are generally diffused with extensive ranges, but this is probably because of the nature of their habitats rather than the size of the genera to which they belong; likewise, the wide distribution of simple plants is also unrelated to genera size. (The reason that simple plants range extensively will be discussed in the chapter on geographic distribution.)
Because I think of species as being just well-defined varieties, I anticipated that species from a large genus would tend to have more varieties than species from a small genus: if many closely related species (i.e., members of a genus) have already formed, many incipient species should still be forming. Where many large trees grow, we expect to find saplings. Where many species within a genus have arisen through variation, circumstances have favored variation, and we expect them to still favor variation. However, if we consider each species a special act of creation, then there is no apparent reason to expect more varieties in a large genus than in a small one.
To test this I arranged the plants of twelve countries and the coleopterous insects of two districts into two groups with, again, species from larger genera on one side and those from smaller genera on the other. The species in the group of large genera invariably have a greater proportion of varieties than species in the group of small genera. Moreover, species of large genera with any varieties invariably have a higher average number of varieties than species of small genera. The same results follow when the division is made after excluding genera with four or fewer species. These observations are understandable if species are just well-defined and permanent varieties, because wherever many species within a genus have been generated, the generation should still be happening, especially because the process of species formation is probably slow. This is in fact the case if varieties are considered incipient species. This is not to say that all large genera vary greatly, their ranks now swelling with species; moreover, some small genera do vary and are increasing. It would be fatal to my theory if this were not so, because geology plainly reveals that small genera have often increased greatly and large genera have peaked, declined, and disappeared.3 All I want to show is that if many species have been formed within a genus, on average, many are still forming – and this is supported by my data.
Again, there is no infallible criterion by which a species can be distinguished from a well-defined variety, and if intermediate links between doubtful forms cannot be found, then naturalists are compelled to come to a determination based on the amount of difference