I was delighted to ignore all the posts by Mediocre Dave, which the Moderator has so kindly identified as worthless. But somewhere in those 3 or 4 pages I noticed that someone else had written that the various "races" of Homo sapiens should be called "species."
We should recognize that the word 'species' has never been precisely defined in biological science.
https://en.wikipedia.org/wiki/Species_problem
Most textbooks speak of species in terms of the ability to mate and produce fertile offspring. That definition was invented (or at least popularized among biologists) by Ernst Mayr in the 1940's. But there are other ways of defining species in use.
Historically species were defined morphologically, by physical appearance, and many species definitions today are still based on that. And there are quite a few morphologically defined species such as different species of bears that are reproductively isolated due to geography but are known to successfully mate when brought together in zoos or wherever. (And hence might be better described as 'subspecies', but are still called 'species' for historical reasons.)
In paleontology, including human physical anthropology, it isn't possible to directly observe mating and its success, so species attributions are typically morphological, based largely on the shapes of fossil bones and other hard parts.
In non-sexual organisms like bacteria that often appear identical to the naked eye (tiny spheres and cylinders) species attributions are typically functional, based on biochemistry and what kind of fermentations the bacteria can perform, or whatever the defining biochemical ability happens to be.
In the quote in the OP Darwin seems to have suggested what might arguably be yet another criterion for species, namely common evolutionary descent. (This is the basis of contemporary cladistic classifications, though it might arguably be more applicable to higher taxa.) As plausible as it sounds, evolutionary descent can't be directly observed and needs to be reconstructed, and those reconstructions are typically based on morphological characteristics.
Homo sapiens IS a species.
I agree with that, while recognizing that the word 'species' isn't precisely defined.
Distinct populations within a species are called "subspecies."
In zoology, 'subspecies' is the only internationally recognized sub-specific taxonomic category. (In botany, there are several.) It's even less clearly defined than 'species', though in zoology it typically refers to recognizable variants that can mate successfully with other variants but typically don't due to geographical separation or whatever.
This vocabulary is used for all living things, with one single exception: We call distinct populations of Homo sapiens "races" instead of "subspecies."
I think that's more of a cultural decision than a biological one. And more recently it's become tremendously politicized. Issues of race have become identified with issues of good and evil, so definitions of 'race' have become tremendously fraught.
I think that recognizable geographic variants in the human species might justifiably be called 'subspecies' based on how that word is used with other animals. These categories do seem to be awfully blurry at the edges of human populations though, where there's increased opportunity for contact with different populations. I expect that we would observe that with animal subspecies too, if they came into contact and intermixed on a large scale. That intermixture of human varieties seems to be increasingly common as world travel and international migrations occur. Geographical reproductive isolation has broken down, at the individual level at least. It might still be more visible in averages over large geographically separated populations though, since mating in a population is still typically with locals who can be expected to be more similar than distant individuals from different populations. So these kind of categories possess a sort of momentum.