I'm about halfway through Keith Thomson's book The Legacy of the Mastodon: The Golden Age of Fossils in America, and so far I have learned quite a bit. Even though the book covers already well-trodden ground (the "American incognitum," Cuvier's mosasaur, Mary Anning, the discovery of Hadrosaurus), Thomson also pays attention to some lesser-known paleontological personas like Ferdinand Vandiveer Hayden and G.W. Featherstonhaugh.
Although Thomson's prose is easily accessible, the book isn't particularly well-written (the short chapters are essentially outlines of particular events or persons in chronological succession), and there are a number of mistakes in the book. Some of these could have been fixed under the watchful eye of an editor, but there was at least one mistake in particular that is of a more pernicious breed.
The specific error I am referring to is on page 83, and concerns the Philadelphia physician and anatomist Richard Harlan (you can see a collection of his scientific work here);
In the same year  Harlan also described a giant fossil sent from Arkansas as a plesiosaur, which he named Basilosaurus (king reptile) cetoides (whalelike), although in fact it really was a whale, so Harlan's science clearly might be called into question. However, he was merely doing his best with the material and information at hand, and the errors that he and other naturalists made give us a view of the difficulties of working with the contemporary state of zoological knowledge.
Even though Thomson qualifies his judgment of Harlan, the passage still seems a bit Whiggish to me. Perhaps if Thomson knew a bit more about the circumstances under which Harlan described Basilosaurus and the circumstances by which it was shown to be a whale he would not be so quick to impugn Harlan's scientific merit.
In 1832 two judges from the south, Judge Bry from Arkansas and Judge John Creagh from Alabama, sent enormous fossil bones found on their respective properties to the American Philosophical Society in Philadelphia. These remains piqued the curiosity of Harlan and fellow Society member J.P. Wetherill, and they asked Creagh to send them any more bones of the enigmatic animal that he could find. Parts of the skull, jaw, limbs, ribs, and backbone of the same kind of animal were quickly dispatched to Philadelphia, and Harlan took up the task of describing them.
From the fragmentary remains he had been sent Harlan supposed that he was studying a huge marine reptile related to the ichthyosaurs and plesiosaurs, although he was not absolutely sure. In his description he wrote;
If future discoveries of the remaining portions of this skeleton, should confirm the indication above pointed out, we may suppose the genus to which it belonged will take the name, not inappropriately, of Basilosaurus.
Indeed, there was at least one character that perplexed Harlan. The teeth in the jaw fragment were different in each position, something characteristic of mammals, not reptiles. While this was certainly vexing, some naturalists across the Atlantic used the apparent existence of a reptile with mammal-like teeth to explain how what seemed to be mammal jaws from Jurassic sediments were really those of reptiles (as, for some, the existence of mammals under the feet of dinosaurs was doubtful). This was the position of H.M.D. de Blainville, but Richard Owen cautioned against using Basilosaurus as proof of anything until it received a better description.
Owen got just that chance in 1839 when Harlan visited England and brought the Basilosaurus fossils with him. With Harlan's permission, Owen studied the jaws and even sliced open some of the teeth to determine the true affinities of the ancient mammal. What he discovered was that Basilosaurus was not just whale-like, it was actually a whale. With Harlan's consent Owen re-named the animal Zeuglodon for its "double-yoked teeth," and the uncertainties were quelled. (An accident a number of years later in a German museum provided further evidence for the mammalian nature of the animal, but that's another story.) The reason that we call this ancient whale Basilosaurus today, though, is that Harlan published first, and even though he acceded to Owen's changed his name still had priority.
Given these facts, I found Thomson's short treatment to be irritating. Working with only a few fragments, Harlan tentatively asserted that the bones he described were from an immense marine reptile. Before the decade was out, however, he got a second opinion from one of the most famous British anatomists and found out his original hypothesis was wrong. Harlan did not resist or protest, but changed his mind based on the evidence. Thomson's treatment does not reflect this, turning an interesting story into an oversimplified bit of textbook cardboard.
As I have said before, the most irritating aspect of such mistakes is how easily they could be avoided. If Thomson actually read Harlan's descriptive paper and followed up on how Basilosaurus was found to be a whale I doubt that he could have made such an error.
I know in writing my own book I will be bound to make a few errors, but I at least try to read every source that I cite. Doing so sometimes provides new information and prevents mistakes, other times it does little but support my original interpretation, but I consider it a point of personal integrity to have actually read the work of people who I am discussing. Otherwise I might as well be making things up, and I certainly don't think we need to keep ingesting and regurgitating the same inaccurate stories.
What's really odd is that a few decades after Owen connected Basilosaurus with the whales, some people considered it to be a pinniped or their sister group! This was undoubtedly a minority opinion, even at the time, but it was a bit of history I've never heard mentioned outside of one source:
(see pages 428 and 554 - look at that "phylogenetic tree"!)
Cameron; I've seen that mentioned in a few places, too. It was the opinion of D'Arcy Wentworth Thomson, but (as you say) not many seemed to subscribe to it. Pinnipeds were used as representatives of the trasitional types whales may have gone through, though, although W.H. Flower preferred a beaver-like ruminant during a lecture on whales in the 1880's.
I think some authors also regarded Basilosaurus as an aquatic creodont (a group that at this time also included mesonychids and other carnivorous "ungulates") in the late 19th century. If I remember correctly, Oskar Fraas proposed that idea, and so did Florentino Ameghino, who thought that the homodont mammals (whales + "edentates" + monotremes) had a reptilian origin separate from the other mammals, so Basilosaurus didn't fit in as an early whale.
Lars; Thanks for the extra information. It seems that after On the Origin was published Basilosaurus was one of the fossil forms that many tried to make sense of in terms of its evolutionary position. Even among those who agreed it was a whale, whether it was on the "main line" of evolution or an "offshoot" was debatable.