Campanula is the largest genus in the family Campanulaceae. Its 300 or so species are confined to the Northern Hemisphere and are found there on all continents except South America, with especially heavy concentrations in the areas of the Caucasus Mountains and in an arc from the Swiss and Italian Alps down through the Balkans. Fewer than a dozen species, notable among them C.s lactiflora, latifolia, latiloba, persicifolia, and medium (Canterbury Bells) are taller plants, staples of the "cottage" and perennial garden. A very few, like C. rapunculus which was grown for its edible root (and which figured in the Grimm's fairy tale "Rapunzel"), or C. rotundifolia, the Bluebell of Scotland, or C. carpatica which is just large and vigorous enough to hold its own at the front of an average herbaceous border, have some familiarity for the gardening public. But the majority are rock garden plants, some seeming to specialize in growing in the most inaccessible mountains, and so are less well known generally. Its common name is Bellflower and many species are suitable for the rock garden. Quite a number of these grow well in Zone 4, and some are grown easily even in Zone 2. Five favorites for Minnesota are C. 'Samantha', C. carpatica 'White Clips', C. medium 'Calycanthema', C. glomerata and C. planiflora (which is actually a Mendelian recessive dwarf of C. persicifolia, but which is often given specific status).
The flower consists of five petals fused at the base forming the corolla, and five leaf-like sepals below called the calyx. The presence or absence, relative size, shape or orientation of additional growths between each sepal called "appendages" are often used by botanists to differentiate between similar species. The shape of the corolla is often described as campanulate (bell-shaped), infundibular (funnel-shaped),bowl-shaped, or tubular when the petals are fused along much of their length, or stellate (star-shaped) when fused only near the base.
The flower color for campanulas in general is overwhelmingly blue, although seldom is it a true blue. The admixture of red pigment results in descriptives such as amethyst, Tyrrhean purple, lilac, violet, and wine. In addition to genetic mutations in some species giving white versions of the standard blooms, usually designated 'Alba', there are some species whose flowers are naturally white or nearly so, a few pinks, and a very few whose color is described as "yellow" (though it is a dull color closer to "straw").
[Unless otherwise noted, the Literature section is the opinion of one reviewer.]
For those wishing to study campanulas in more detail there are essentially three books to choose from: "Campanulas" by Crook, 1951, "Campanulas" by Lewis and Lynch, 1998, and "Dwarf Campanulas" by Nicholls, 2006.
Lewis & Lynch is, in some ways, "Campanulas Lite" - thinner, less text, less technical, far fewer species, larger photos, and more tall border-variety species of limited interest to rock gardeners. But the photos are excellent, text is clear, many cultivars are detailed, and the intro includes instructive diagrams showing cutaway views of flower structure. If your interest runs to plants taller than 6 in./15 cm. this book provides information not available in the others. Not an essential textbook, but good enough that it is referenced by Nicholls in his book, which is.
For 65 years Crook was the standard, and it - and he - are undeniably classics, however flawed they might be. After all this time it is no surprise that a good deal of the information is obsolete. Some species are no longer considered to be species, some have changed names, others have been reassigned to other genera, new ones have been discovered, . Even at the time it was written much of the info was "of botanical interest only" rather than of use to the average - or even specialist - gardener. A considerable amount of space is given to naming and describing a species only to dismiss it as being probably not really a campanula, or only a size variation of another species, or lost to science, or a synonym, not to mention the ones which are impossible to propagate or ineradicable weeds (which I freely admit should be listed).
There are a number of jarring instances of self-contradiction or lapses in logic, and it's surprising they were not corrected in subsequent editions. For example, in discussing the validity of the many names put forth as distinct species for those plants which differ in some slight way from C. rotundifolia , he grudgingly allows that only one, C. linifolia, is just different enough to warrent specific status. However, under C. jaubertiana he opens by saying it "appears to come between C. linifolia and C. rotundifolia," species he's told us are almost identical. Two paragraphs later, he casually mentions that jaubertiana only differs from C. stolonifera, growing in the same area, by having its stolons all on the surface. (Neither rotundifolia nor linifolia are stoloniferous, by the way.) Looking up stolonifera, there is no mention of relationships to any of these three species, but he says it much resembles C. cochleariifolia, a completely different species! (Intriguingly, stolonifera is no longer considered to be a species.) So Crook was claiming that the stoloniferous jaubertiana was midway between two nearly identical (to each other) non-stoloniferous species, and almost identical with a different, now unidentified species that looked like a fourth, unrelated species!
In addition he suffers, as does Nicholls, from inability to cross-reference,and the absence of an index means you'll have to make your own notes in the margins. Photos are in b&w, and in the 1951 edition they are glossy and sharp, in the 1971 reprint matte and fuzzy. Other editions unknown.
On the other hand, it is a classic, half the information is still relevant, and for cross-referencing it is invaluable.
Nicholls has, of course, superseded Crook, if for no other reasons than for the color photos and up-to-date info. But there are other reasons; he includes a good number of hybrids and a sizable section of closely related genera of considerable interest to the rock gardener. He, too, includes many obscure and probably unobtainable species (then again, where does one draw the line?), but not the crypto-species and useless tidbits Crook does. There's lots of cultural and anecdotal info, obvious experience, but more than a few instances where you'll search in vain for any indication of whether the species in question wants alkaline soil or full sun or lots of water or whatever, because every preference is not given in the description for every species. And the relationship among species is not always given either, so you can't easily search among related species to try to extrapolate.
He, too, has a few stumbles in logic, or perhaps just unfortunate wording. For example, under C. jaubertiana (maybe the species is jinxed) he says it grows in northern Spain alongside C. cochlearifolia, "but they never appear to hybridise." The next paragraph concerns a plant that had been found and tentatively identified as a subspecies of jaubertiana, but which he says is probably a natural hybrid with cochlearifolia! If there had been a qualifier, or the "...never appear..." had instead read "...seldom appear...", it would have been an interesting anecdote rather than a self-contradiction. It is not, unfortunately, an isolated example of its kind, but is more than made up for by excellent photographs, the inclusion of just about any dwarf species (and related genera) you're likely to want to grow, and an index to keep track of it all. He is, of course, a world authority on campanulas, and awkward phrasing or unsatisfactory cross-referencing are, I suppose, more a problem of editing and proofreading than content.
So, if you're going to have only one campanula book, Nicholls is the one. If you already have Crook, you need to get Nicholls, too. And if you're really serious, it wouldn't hurt to get L&L as well, just on the off chance that a tiny bit of info that you need won't be in the others, and you'll be able to find it in there.
In the genus there are groups of related species which have been traditionally recognized. These have no formal standing in the way that, say, porophyllum saxiphrages or bearded irises have, and group membership - or even existence - is debatable. In a section at the end of his book Crook recognized seven groups, but that number is misleading.
Two of the groups, lyrata and rupestris, consist of difficult Grecian monocarps, between which he described no convincing differences, and which are of limited interest to the average gardener in any case.
He lists a carpatica group, then dismisses the many proposed member species as mere varieties of the type, mentioning only two by name and by clever wording managing to present them without vouching for their legitimacy.
For the rotundifolia group he also questions whether the proposed species are just understandable variations in this circumboreal species. Elsewhere in the text, he makes it clear he accepts only linifolia as a separate species, which hardly constitutes a "group."
Another group of limited general interest is the only one not to be named for its best-known species, the Himalayan group, which according to Crook consists of unattractive weedy annuals, although Lewis & Lynch are fond of cashmeriana. It seems at least some of the species are perennial, but so undependable that they must be treated as annuals and resown each year with saved seed, not an easy task as the seed is exceedingly fine.
The garganica group, with long, usually lax or ascending stems covered in small infundibular or stellate flowers, is centered around the Adriatic. Besides garganica itself, the group includes poscharskyana, portenschlagiana, fenestrellata and a dozen less well-known species and possible sub-species and varieties such as elatinoides and istriaca. C. elatines, the only calcifuge among the otherwise lime-loving species which superficially resemble garganica, has recently been separated (according to some scientists) from the group based on genetic studies, while C. reatina, discovered only a decade ago and confined to some cliffs in Lazio in central Italy, has been added by the same method.
Perhaps the most controversial is the tridentata group. Centered in the Caucasus, its members typically have thick taproots, overwintering buds nestled in brown rosettes of spoon-shaped leaves, thin ascending stems to about 6" and solitary large cup-shaped blue flowers. All sources seem to be required to quote Ingwerson in describing "the confusing group...all grading into each other to a bewildering extent" as explanation for proposing as many as 20 species for inclusion.Tridentata, saxifraga, bellidifolia, and aucheri are the core species everyone can agree on, with anomala, ardonensis, kryophila, ciliata, and biebersteiniana being among the runners-up.
The betulifolia group is one which Crook did not know, as only its namesake member had been discovered in his time. It is a small and coherent group, easily distinguishable from other campanulasand so not controversial. Its members, betulifolia, trogerae, seraglio, choruhensis, and kirpicznikovii are found in the Caucasus and southward to Eastern Turkey. From large rosettes of birch-like leaves, ascending or lax stems hold large wide-open whitish blooms. In all except kirpicznikovii there is a pinkish tendency; buds may be pink or even wine-red, some forms will have pink flowers. Kirpicznikovii is described as having yellowish buds and creamy white flowers.