At

completely opposite inference from the same premises. Suppose, for instance, (and I state the incident as it was related in the Courant newspaper, a few years ago), a large trout to be caught in the Canal, in a nearly exhausted state, with a frog mounted upon its back, and that the said trout was injured in the neck, opposite the frog's mouth,-might not the learned doctor, like the narrator of the circumstance in the newspaper, very naturally suppose that the frog was in the act of devouring the trout? And, supposing the breed of frogs to be of national importance, and their food a grave matter of philosophical enquiry, would not this fact be considered as incontestable proof of the nature of their aliment, however different from the frog's habits, and direct evidence that the reptile had taken at least one mouthful? The fact of the frog on the back of the trout, and clasping it with its arms, is of undoubted occurrencethe reason assigned is mere matter of opinion, and in this case would be perfectly erroneous. To those acquainted with the natural history of the frog the solution is apparent, without involving the crime of trout-murder. the usual period of the year, the instinct of reproduction in these animals is strong, and, failing females of their own class, the male frog frequently sits the usual time upon the back of a fish. (See Blumenbach and Spallanzani.) I have heard of ponds in England being nearly cleared of trout from this cause, where frogs abounded; the trout being literally ridden to death by these amphibious equestrians. Now, the natural conclusion of one ignorant of the habits of the animal, on such an occurrence coming under his notice, would be that the frogs had seized upon the trouts for the purpose of devouring them; and "one more imaginative might naturally enough conclude that imps in the shape of frogs were running sweepstakes in a submerged racing-course. It need not be said how far from truth these inferences would be; but such is the mode in which the author of the paper before us treats the evidence given by professional fishers, and others, before a Committee of Parliament. If they err in opinion, their evidence as to facts coming under their express cognisance is not to be believed.

Mr Alexander Fraser, a salmon

curer, who published a Natural History of the Salmon in 1830, is attacked in the same manner; and his statements, founded on experience, are treated with disrespect, and as not to be believed, because he counted only 54 vertebræ in the backbone of the salmon, when, according to Dr Knox, there are really 61. Thus," says the Doctor, "an experience of forty or fifty years as a salmon-curer and catcher has not enabled him to count the backbones correctly."-(P. 501.) And again, because Mr Fraser has omitted to mention the "internal parasitical animals," "I confess," says Dr Knox, "this excites strong doubts in my mind as to the accuracy of Mr Fraser's observations generally, and causes me to undervalue altogether his forty years' experience as a salmoncurer." (P. 502.) And again," Mr Fraser has a mind capable of rising above all prejudices in the support of truth."—(P. 502.)

Now, if to be able to count the number of vertebræ in the backbone of a salmon correctly be the rule of judging of the credibility of testimony in other matters, what shall be said of Dr Knox, if his enumeration be not itself correct? Mr Yarrell, whose knowledge of fishes, internally as well as externally, requires no praise of mine, makes the number of vertebræ in the salmon sixty and so does the Rev. Mr Jenyns, in his accurate work on the British Vertebrate Animals. "According to Dr Richardson" (says Mr Yarrell), "the cœcal appendages are in number from 63 to 68; and several observers have stated the number of vertebræ to be sixty, which I have repeatedly found to be correct."British Fishes, ii. 6.

But Mr Fraser, it seems, has given other and greater offence in speaking of the salmon's food. "In respect to the food of the salmon" (says Dr Knox), "Mr Fraser has notions also perfectly local; that is, confined to himself. And as the whole passage admits of no sort of analysis, and as, indeed, no person having the smallest knowledge of natural objects would think it necessary to read the article twice, much less to examine it seriously, we shall simply quote his own words, and then leave it :— Their digestion is so quick, that in a few hours not a bone is to be discovered. Of this I have had various proofs, in trouts caught by a par as a bait in set

lines.

Fire or water could not consume them quicker.'"-(P. 501.) Sir Humphry Davy shares in the contemptuous reprehension implied in the introduction to this paragraph; for he says, "their digestion appears to be very quick"-(Salmonia, p. 130); and other writers corroborate the observation.

As to the food of the salmon, then, it appears clearly that this food, both in rivers and in the sea, was well known, and recorded by almost all writers on the natural history of fishes, long before the observations of Dr Knox appeared. That food is, gene. rally speaking, worms, insects, and small fishes; the first term including the Echinodermata of modern writers, and the second the modern class Crustacea. The author's criticism on Dr Fleming saying that salmon go "into estuaries in search of worms and other bait," is almost unworthy of notice. Worms may mean only earth-worms in the vocabulary of Dr Knox, though the use of the word in the plural number might have suggested to a person so learned, that there might be marineworms as well as earth-worms in the estuary alluded to. But in point of fact, one of the stomachs now on the table, confirms even the verbal accuracy of Dr Fleming, had he even meant, as is sneeringly imputed to him, merely earth-worms-for there is actually an earth-worm in that stomach, washed down probably from the banks of the river by the receding tide. The evidence of practical fishermen and others fully establishes, what previous writers had asserted, that small fishes, particularly sand-eels and shrimps, form a chief portion of the food of the salmon, without, however, excluding worms, and other animals found on the shores which salmon frequent. That they may also feed on the ova of the Asterias glacialis; on the ova of fishes; and even, like the haddock, swallow this and other species of Asterias entire, I have no reason to doubt, and would willingly admit, even on less than the single evidence of Dr Knox, because that class of animals is, amongst others, stated by all authors to form the food of the salmon. And the existence of ova in the intestinal canal or stomach of the salmon and herring, when the other portions of the food are decomposed, is easily accounted for, from the known resistance of the coriaceous envelope of

the ova of fishes to the action of the gastric fluid. But beyond this single remark, that ova of the Asterias glacialis is found occasionally in the stomach of the salmon when in season, there is nothing in the memoir that can be said to have extended our knowledge of the food or natural history and habits of the salmon.

The next portion of Dr Knox's paper which claims notice is "the Generation of the Salmon, the Growth and Progress of the Smolt, and the descent of the kelt or spawned fish to the ocean," (P. 471); and here again the author seems to labour under a lamentable ignorance of what has been recorded on this subject before the appearance of his paper; for he declares he knows of no continued series of observations on the subject, published by any one, of an authentic nature, and so as to admit of no doubt."" To fill up this chasm, he resolves to detail the history of the salmon smolt, from its first deposition under gravel, in the form of an egg, to its ultimate disappearance from the fresh water streams; "remarking, that every thing stated therein fell under my own immediate personal observation." The dates of observation are- - Nov. 2 (1832?), Feb. 25, March 23, April 1 and 19, and May 5 (1833?). The results of these observations will be stated in the sequel of this notice, as compared with the prior observations of others. It may now, however, be mentioned generally, that Dr Knox has not stated a single fact regarding the deposition or growth of the ova of the salmon-the periods of their ascending the rivers where they breed, and their return again to the sea,which had not been observed and recorded with much greater minutene: s prior to the publication of his paper.

Salmon ascend the British rivers at different periods according to the seasons, generally from September to January, and deposit their spawn during the months of November, December, and January. This is fully ascertained by the evidence led before the Committee of the House of Commons. The names of the witnesses need not be here mentioned; but they state the period of salmon ascending the rivers on observations for periods varying from a few years up to no less than forty. This evidence was taken in 1824-25. Dr Knox, in the single

instance he mentions, says, that a pair of salmon were observed in the Whittader, one of the tributaries of the Tweed, to be spawning on Nov. 2. The ova remains in the spawningbed or gravel for three or four months, according to Dr Fleming; according to Mr John Johnstone, from the ova deposited in November, December, and January, the young rise from the gravel in March, April, and May; according to Mr John Halliday, the spawn deposited in November, December, and the beginning of January, is disengaged from the spawning-beds from 10th March to 10th April; so that it appears, on an average of seasons, the salmon roe lies about four months, or 120 days, in the gravel beds before the young appear. But according to Dr Knox, in his single observation of the Whittader pair of salmon, the ova took 142 days "to become fishes somewhat less than an inch in length," but still "embedded in the gravel." (P. 473.) On the 19th of April the fry are "eight and even nine inches long ;" and on the 2d May they still abound in the tributary streams, but are not so numerous as before; they are not increased in size, and are, in all probability, the fry of a later deposit."-(P. 473.) So that the amount of Dr Knox's information here is, that the spawn of a single pair was hatched in April, and other families of other fishes were of a later deposit, and appeared in May. The witnesses examined before the Committee of the House of Commons had stated all this much more fully in 1824 and 1825. Thus, Sir Henry Fane says the fry descend in April and May Alexander Fraser, early in April and May-Rev. Dr Fleming, March, April, and May Mr George Hogarth, jun., April and May-Mr William Stephen, March and April, to 14th May, according to the temperature of the season and the situation of the different rivers. It appears, therefore, that Dr Knox's single observation is corroborative of the evidence led before the Committee, as far as a single instance in a single river can; only it is to be observed, that he makes his single observation, made he says by himself, the rule for spawning in all rivers by all salmon, without reference to season or situation; while the evidence of practical men give the average result of many years of observation, extended over every variety of season,

and over all the rivers of the empire.

The mode in which the salmon deposit their ova in the gravelly bottom of shallow streams, is minutely described by Mr George Little (Report, p. 108-9), and, indeed, has been known for centuries; for John Monipennie has so long ago graphically recorded the manner in which this instinctive work is done. "In harvest," says he, "they come from the seas up in small rivers, where the waters are most shallow, and there the male and female, rubbing their bellies or wombs one against the other, they shed their spawne, which forthwith they cover with sand and gravel, and so depart away."-P. 195.

As to the developement of the ova under Dr Knox's "own immediate personal observation,"—though the ova would not, it appears, bear transmission to Edinburgh-it really seems unnecessary to notice such an evidently imperfect experiment, particularly since a very full and interesting account of the gradual developement of the ova of the salmon, accompanied by an accurate engraving, is given in the evidence of a gentleman before the Parliamentary committee. To that engraving, and the description of the report in general, I beg to refer any one who takes an interest in the subject. At the same time it is proper to mention, that there is an interesting article on the "Spawn of Salmon," by Mr Schonberg, printed in Sir David Brewster's Journal of Science in 1826, accompanied also by an engraving of the ova in different stages of growth. Both these sets of figures, and the accompanying details, correspond with one another in every essential particular; but both at the same time differ widely from the details given by Dr Knox. Neither does the Doctor even hint in his paper at the existence of such details or figures, though he could scarcely be ignorant of what is stated in the Report, which, he asserts, he had repeatedly read over.

To pass over discrepancies which materially lessen the value of his remarks, Dr Knox asserts, that "ova taken from the bed of a river at any time from January to March inclusive, and not shaken or carried far, will live and become developed, i. e. grow to fish of about an inch in length in a small glass full of water, changed not oftener than once a week,” p. 476.

Then follows a passage in which temperature is said to have some effect in hastening or retarding the developement of the ova, though in his reckoning by days such agency is necessarily excluded; and he adds, that after having cast the slough, they will live about ten days (seldom or never longer) in water unchanged, apparently thriving, growing, and darkening in colour (if exposed to the light) every day.”—P. 477.

This assertion, of the ova and salmon fry living a week and even ten days in a small glass of unchanged water (almost the only original observation in the paper) is in complete contradiction to all experiments that have been made on the developement of the ova of this genus of fishes. "It is said by Sir H. Davy" (says Dr Knox), "on the authority of a person of the name of Jacobi, whose writings I have not met with, that the ova of salmon are deposited in the gravel of rivers under streams, in order that they may be perfectly aërated, or exposed to water which is so. This reason, which appears so plausible, is probably not the true one.”—(P.476). The person of the name of Jacobi here mentioned, though unknown to Dr Knox by his writings, was a Counsellor of State to the King of Prussia, and a well-known experimenter on the artificial fecundation of the ova of fishes. His experiments appeared in the Berlin Transactions for 1765, and have been referred to with approbation by almost every writer on the subject of fishes since. These experiments were made chiefly upon the ova of the genus Salmo; and he found that by expressing the unimpregnated ova in water, and afterwards applying the milt, the ova became impregnated, and went through the usual developement. In making these experiments, one thing essential to their success was found to be necessary, and this was the frequent, almost incessant, changing of the water; and hence he justly concluded, that the aëration of the water where ova are deposited, is necessary to the developement of the

tity of trout in rivers or lakes." ́ (Salmonia, p. 82). "In all experiments of this kind" (continues Sir Humphry) "the great principle is, to have a constant current of fresh and aërated water running over the eggs. The uniform supply of air to the fœtus in the egg is essential to life and growth; and such eggs as are not supplied with water saturated with air are unproductive."-Salmonia, p. 82, 83.

This necessary aëration, and exposure to the influence of the sun's rays, explains at once why salmon seek the gravelly bottom of shallow streams for the purpose of spawning; and the same instinctive impulse which guides the salmon, induces the herring and the cod, among numerous other fishes, to approach banks and shores, and thus carry boundless provision to countless animals. It was a curious circumstance in Jacobi's experiments, that the effect of his impregnation of the ova with the milt, often produced in the trout monsters with two heads, &c.-so different are the rude attempts of man from the instinctive workings of nature.

The experiments on the salmon ova by Mr Hogarth and Mr Schonberg, who both traced their developement, from the first appearance of life, till the animal was an inch in length, further demonstrate the necessity of this aëration. With "frequent changes of water, Mr Hogarth succeeded in hatching the ova, and by changing the water frequently' the animals appeared vigorous for three weeks, after which they became restless and uneasy.' (P.92). Mr Hogarth also tried one of the fry hatched in fresh water, if it would live in salt water; but found that it immediately showed symptoms of uneasiness, and died in a few hours."" (P. 92). The figures of the ova, and the young animal in its different states, were drawn by an artist, at the request of Mr Hogarth, and an engraving of them is appended to the Report of the Committee on the Salmon Fisheries.

Mr Schonberg found the frequent change of water equally indispensable. "Changing of the water" (says he), "and if possible from the same river, must be repeated hourly, and they must likewise be exposed to the sun's influence." (Journal of Science, v. 238). The developement of the ova is well represented in the engraving

which accompanies Mr Schonberg's valuable paper. The details of his experiments are more extended than those of Mr Hogarth; but both agree in all the more important points.

Dr Knox's experiment, although said to have been made under his own eye, is contradictory of the fact that aëration of the water is necessary, as he, or the person who took charge for him, appears to have kept the fry in water unchanged. But better evidence than this will require to be produced before we can give up the hourly and daily observations, bearing all the marks of truth, made by Mr Hogarth and Mr Schonberg, in opposition to the statements and examinations of Dr Knox or his assistants. Of Dr Knox's candour and fairness in not referring to the experiments of those gentlemen, though one of them appeared in the Parliamentary Report which he so much abuses, and the other in a Journal consulted by every one with any pretensions to science, I leave others to draw the inferenceslimiting myself to the plain statement of facts. Of course I hold, with all writers on the subject, except the author of this memoir, that the aëration of the ova by the frequent change of water is necessary to the developement of the salmon fry in rivers; and that this, and a certain exposure to the rays of the sun, influence the approach of fishes to the banks and shores upon which they deposit their spawn.

In reference to Jacobi's experiments, the stocking of ponds or lakes with any desired species of fresh-water fishes, is, by these experiments, proved to be comparatively easy; for he found that the ova could be impregnated, and the animals from these ova hatched, after the parent fishes had been dead four days. Even the Vendace of Lochmaben might thus be introduced into other lakes without much danger of failure, by catching a few of these fishes previous to spawning. It is well known that the Chinese stock ponds with impregnated spawn of fishes.

The period of the salmon fry rising from their gravelly bed has been already stated generally as occurring in March, April, and part of May; but this of course depends upon the season. Mr George Little gives decided evidence as to this point. "A great deal" (says he)" depends upon the season at the time of the year,

whether we have an early spring or not; sometimes there may be two or three weeks of difference, according to the season." "I have observed, when we have early warm weather, the fry come early, and when we have a late spring, it is later before the fry rise from the gravel bed." (P. 109).

The descent of the fry to the ocean is, in the Avon, according to Sir Henry Fane, in April and May-in the Ness, according to Alexander Fraser, early in April and May-in the Don, March and April, to the middle of May-in the Dee, April and Mayin the Tay, March, April, and May; and so on, according to the season. The kelts, or spawned fish, descend with the winter and spring floods. But the dates given in evidence by the numerous and respectable witnesses examined before the committee are not be taken as absolute periods, common to every year. The tempe

rature of the season must be a powerful element in determining the ascent of the salmon, the deposition of the ova, and the hatching of the ova; in fact, the temperature and other circumstances, there is every reason to believe, might have the effect of hastening or delaying the process of reproduction, as the same meteorological agents are known to hasten or retard the annual harvest, or prematurely bring out or delay the appearance of many of the insect tribes. With the exception of Dr Knox fixing a determinate period for the developement of the ova in the gravel till the appearance of the smolt, I say, with this exception alone, any reader of his paper, and the minutes of evidence, might have naturally enough supposed, that, with regard to these points, he took his information from the Parliamentary Report, which he reprobates, and from the testimony of witnesses whom he declares unworthy of all belief; and the strong coincidence between the Doctor's periods of migration as related in the Transactions, and what was stated by these gentlemen six years before, must either appear very strange, or the witnesses have not deserved that unmannerly abuse which has been dealt out to them under the sanction of the Royal Society of Edinburgh.

Another particular noticed by Dr Knox in his observations upon the salmon smolts is, that they will not "bear the slightest handling-they