My dear Darwin,

Pray do not apologize for the length of your letters on nomenclature as I am always glad to meet with any one who appreciates the importance of the subject, not indeed as being a science, but the vehicle of science, in which if its wheels be ricketty, Science herself will be upset,—or run away with.

A few years ago, before the law of priority was so generally admitted as now, the vanity of species-mongers took a different turn, each man trying to supersede his neighbour by “improving” the existing nomenclature, and deliberately changing names however generally current. (Leach, Wagler,¹ Swainson, Neville, Wood,² &c to wit). To remedy this evil, which was fast chaotizing science the law of priority was loudly called for; & generally enforced. It now appears that vanity is not checked, but only diverted to a new course. Those who once obliterated the claims of others by substituting their own “improved” names (with the accompanying mihi), now hurry to be first in the species-market, like the Covent-gardeners³ with their green peas, a guinea a pint, at Xmas, and unripe strawberries, a shilling a dozen, in April.— Truly a pitiful competition, still It promotes horticulture, and so the vanity of species making promotes the collecting of materials, & to a certain extent the observation of facts so that it is at least one degree better than the improved-name-vanity of which I first spoke. If however there is the least chance of developing this silly vanity into an honourable ambition for the advancement of knowledge, by means of your proposed plan of quoting good Books in preference to first descriptions, I am sure I have no objection to it, and if you like to try the plan in your Cirriped work, I shall look with much interest to see how the plan answers.—

Still this must not be allowed to affect the law of priority. Every Systematic work ought, as a matter of history to place the earliest description first in the list of synonymes, and as a matter of justice, to adopt the earliest specific or binomial name. This is a principle which I cannot bring myself to give up on ground of convenience, especially as by a general adherence to it, convenience would ultimately coincide with justice.

In Gray’s Genera of birds⁴ this principle has been rigorously adhered to, and though I have had to unlearn a great number of specific names in consequence, yet I feel satisfied that the advantages of a simple and certain rule more than counterbalance the inconvenience of its introduction. For instance, Buffon⁵ described and figured some hundreds of new birds under French names only. Gmelin⁶ gave Latin names & definitions to them, & these names have been used by Latham,⁷ Shaw,⁸ Temminck,⁹ and numerous later authors. But (in an unlucky hour I admit) the diligent G. R. Gray dug up in the Library of the Brit. Mus. a very rare tract by a Dutchman named Boddaert,¹⁰ containing a set of Latin binomial names, referring (without a word of description) to Buffon’s species, & 5 years prior to Gmelin. Most unfortunate I grant it is that ever Boddaert lived or Gray discovered his book, still Buffon’s descriptions are good & recognizable; Boddaert had as good a right as Gmelin to give Latin names to Buffon’s species; it was Boddaerts misfortune that his work remained in obscurity while Gmelin’s became popular; & I cannot therefore do otherwise than follow G. R. Gray in using Boddaert’s names, & can only exhort others, for the sake of future peace & tranquillity, to do the same—

I had intended (as you advise) simply answering your queries without launching into argument, but I could not stop when once started.

So now for case A.¹¹ Lister, & Hill’s first edition are prior to 1758, & are therefore altogether out of court. I should say the same of Hill’s 2^d edⁿ. if it be a mere reprint, but if it be an improved edition, revised by the author himself, and adopting the binomial method, it would be entitled to the same privileges as an original post-Linnean work— In that case Pentalasmis would be the name to adopt, but if Hill’s 2^d edition be only a reprint, then Anatifa, Bruguière, 1792 would stand (= Pentalasmis, Leach, & = Pentalepas Blainv. as synonymes). Or if there is reason to suppose that Anatifa was a mere blunder of Bruguière or his printer, of course you may write it Anatifera, & still quote Bruguière as the author.¹²

B. As you state the case, Balanus must (however uncomfortable the change may be) give place to Balanites, and I can honestly say that I for one would always use the latter name, which is as much as I need say at present.¹³ But it appears to me that both Balanus and Balanites must yield to the old Linnæan name Lepas (restricted quantum Inf.) the type of which is Lepas balanus.¹⁴ The restriction of a genus does not cancel the original name. (See Brit. Assoc. Code, [SYMBOL] 3. and 4.)

Of course the law of priority can only be applied as far as there is evidence as to date. In works coming out in Nos. I always write the month and year on each sheet as I get them. Temminck’s Planches Coloriées took 18 years to publish 600. plates,¹⁵ but I never could get the exact date of each livraison. I have therefore been obliged to assume that it was published with uniform rapidity, and by dividing 600 by 18, I get approximately the date of each plate & act accordingly with his names.—¹⁶

C. In regard to Leach’s Acamptosomata and Campylosomata I may say once for all that I never contemplated extending the law of priority, to higher groups than Genera and Species. I remember I wished to state this distinctly in [SYMBOL] 1, but it was overruled, though the general tenor of the code sufficiently shows it. The names of higher groups than genera, must I think be continually modified with the progress of science.

D A recognizable plate is per se regarded as a sufficient (tho’ not commendable) specific definition, & it therefore, if accompanied by a binomial name, acquires the same right of priority as a description would do (e.g. plates in Gray’s Genera of Birds) Of course a plate ought (morally speaking) to be accompanied by a description, still I think it may be conceded that a good plate, duly published, has as good a claim for priority as many an existing attempt at verbal definition.

I shall always be happy to send you an opinion (if worth having) on any case w^h. you may wish to put, & am your dev | H E S.