Popular Science Monthly/Volume 61/July 1902/Studies in the Natural History of the Sacramento Salmon
|STUDIES IN THE NATURAL HISTORY OF THE SACRAMENTO SALMON.|
ASSISTANT, U. S. FISH COMMISSION.
THE following notes are derived mainly from a series of investigations carried on under the direction of the U. S. Commissioner of Fish and Fisheries, by whose permission they are here used.
The Pacific salmons, the genus Oncorhynchus, of which there are five species, are very distinct from the Atlantic salmon of the genus Salmo. In fact, they have no more right to the name salmon than wolves have to the name fox. Anatomically the two genera do not differ greatly—Oncorhynchus having 14 or more rays in the anal fin, and Salmo 10 to 12—but physiologically and in life-history they differ in a marked degree.
The most important difference lies in the fact that the Pacific salmon, of whatever species, dies immediately after spawning once. This is true of both males and females, and is a very remarkable characteristic. It is often denied upon a priori reasoning, the common argument being that, if they all died on the spawning beds the rivers would be full of dead salmon floating down stream. But the common idea that dead fishes float is erroneous. Those that die a natural death do not float, and the salmon is not an exception.
Under natural conditions, the female salmon extrudes the ova, a few at a time, in a swift current near the bottom of the stream. Many are carried several feet, or even yards, down stream by the current, and are nearly always devoured by other fishes, such as the trout, sculpin, Sacramento pike and split-tail. Some of them, at least, are caught in the eddies formed among the cobble stones, and are held near the 'nest,' or hillock of gravel thrown up by the spawning fishes. These
are covered by the sand and protected, though many are covered too deeply and are killed.
Salmon ova are about .25 inch in diameter. The 'shell' is membranous and finely porous, with a minute aperture, somewhat larger than the pores, known as the micropyle. When first extruded the ova
are soft and compressible, and it is while in this condition that they must be fertilized. Within two or three minutes after being deposited in the water they become turgid, and are then incapable of fertilization.
Immediately after the extrusion of a few ova, the female moves away from the 'nest' and the male takes her exact position, or sometimes a little down stream from it, and extrudes a small quantity of milt. The milt is a milky white fluid as full of spermatozoa as blood is of corpuscles. It rapidly disseminates through the water, and is carried away by the current just as the ova were. Doubtless many hundred spermatozoa come in contact with each ovum, though probably only one finds its way through the micropyle, which causes fertilization. The salmon egg is too large to permit a microscopic examination of the process of fertilization, though without doubt it is the same as in other fishes.
Fertilization is secured at fish culture stations by expressing the ova and milt simultaneously into a pan and thoroughly mixing them by stirring with a feather or the fingers. Two methods of procedure are in vogue among salmon culturists. The one, known as the 'dry' method, consists in expressing the ova and milt into a pan that has been merely rinsed with water; in the other, known locally as the 'wet' method, about a pint of water is placed in the pan before the spawning. A careful comparison of the results has failed to show any difference. The 'wet' method requires a less quantity of milt, which is sometimes a desideratum.
Numerous experiments were performed testing the vitality of ova and milt under various conditions, of which I note the following:
A quantity of the ordinary creek water, such as was used in artificial propagation, was spermatized, and a portion of it used for fertilizing ova at various periods after the spermatization, in order to test the vitality of spermatozoa in water. The following results were obtained from one experiment, which may be considered typical:
|Time milt had
been in water.
|Time milt had
been in water.
Similar experiments were performed with the ova, to determine their susceptibility to fertilization after being immersed in water for various periods. The following results were obtained from one series, which are typical of the others:
|Time ova had
been in water.
|Time ova had been
|2||57||More than 7||0|
The average salmon produces about 6,000 ova. In artificial propagation from 5 to 20 per cent, cannot be spawned. These can be removed only by slitting the fish open, in which case a considerable quantity of blood is mixed with them. If the blood is not washed off, the fertilization cannot be made complete, as many of the ova become surrounded with clotted blood. But, if the blood is washed off with water, the ova immediately become turgid and not susceptible of fertilization. The difficulty was overcome as follows:
Among the many experiments testing the vitality of the ova and their susceptibility to fertilization under various conditions, one determined their reaction to normal salt solution. By normal salt solution, is meant water of the same degree of saltness as the body fluids, which in the case of the salmon we assumed to be .75 of 1 per cent., no chemical determinations having been made. Ova were spawned into a pan containing normal salt solution, and after various periods of time a few were removed and spermatized in the ordinary manner, with the following results:
|Time ova had been
in salt solution.
|Time ova had been in
Compare this table with the one showing the effect of pure water on the ova. Acting upon the information gained in this experiment, we removed the unspawned ova from the fishes, washed off the blood with normal salt solution, and fertilized them just as we did other ova, thus saving a large percentage that had previously been lost.
Habits of the Young.
Before this series of investigations was begun, fish culturists were able to hatch artificially from 80 to 90 per cent, of the eggs taken, and it did not seem that much improvement in that line was probable. But there was considerable question as to the best methods of planting the young. We were entirely ignorant concerning the life of the young in the streams, knew nothing of their food, nor of their enemies, and knew only in a general way that they migrate to salt water. Our most important study, therefore, was that of the natural history of the young, and later of the adult also. The following are some of the results of our investigations:
|An Alevin.||A Fry, at the beginning of the Migration.|
The time required for salmon eggs to hatch is about 50 days, though it varies from one to six months, according to the temperature of the water. When the young first leaves the shell, it is attached to a large mass of yolk, and is known as an alevin. It is a very helpless creature, cannot swim and fortunately does not need to eat, the yolk supplying the needs of growth.
For three or four weeks the alevin lies quietly at the bottom of the stream in the crevices of the stones. By that time the quantity of the yolk becomes so small that there is a desire for more food, and the alevin occasionally leaves the bottom to snap at some floating particle. It is at this time, while the movements are slow on account of the unabsorbed yolk, that the young under natural conditions are in the greatest danger from other fishes. In artificial propagation they are protected during this period, and it is only after the complete absorption of the yolk that they are liberated in the streams. At this age they are known as fry.
The fry are practically without enemies. The stomachs of more than a thousand trout taken in streams inhabited by young salmon have been examined, but in no instance has a fry salmon been found, though alevins were common enough. Many Sacramento pike and striped bass have also been examined with a like result. There can be little doubt that the salmon fry in fresh water is able to take care of itself.
Our most extended observations were made on the migratory habits of the fry, and I give a somewhat detailed account of them. In May and June, 1898, we visited all parts of the Sacramento River, from its source to Suisun and San Pablo bays, and even traveled 250 miles down the river in a rowboat. We were equipped with fine-meshed seines which we used wherever it seemed practicable or desirable to gain information concerning the young salmon. We found them abundant everywhere above the middle portion of the river, and in a decreasing number all the way down to the mouth, and along the shores of the bays. We considered them abundant when we caught anywhere from 25 to 400 at a single haul of a 50-foot seine. They were about two inches long wherever taken. The same observations were made again in July. In the headwaters, at the Sims, the fry were as abundant as at the previous examination. There were fewer at Redding, very few at the mouth of Battle Creek, and none at all below the latter point. All that we had found on the previous examination had gone down stream and had passed into salt water. As we afterward learned, the salmon fry observed during this first examination were merely the last of the season's migration, and not all of it, as we first supposed.
While at Battle Creek Hatchery during October and November, 1898, we continued the observations by setting a trap in Battle Creek, and so arranging it that it caught only such fishes as were going down stream. By this means we soon learned that the fry begin their migration much earlier and younger than our previous summer's work had led us to suppose. The following is a record of the daily catch of the trap:
|"||12||Night||17||"||17||8-9 P. M.||17|
|"||13||Day||0||"||21||3-4 A. M.||83|
|Night||7||"||22||4-5 A. M.||6|
|"||14||Day||0||"||24||12-1 A. M.||1|
|Night||10||"||25||1-2 A. M.||5|
|"||16||Day||0||"||26||1-2 A. M.||15|
|Night||8||"||27||2-3 A. M.||0|
|"||17||Day||0||"||28||2-3 P. M.||0|
|Night||6||"||30||1-2 A. M.||49|
|"||21||Night||11||8-9 A. M.||24|
|"||25||5-9 P. M.||4|
Of the 339 young salmon taken in the trap, 322 were under 2 inches long, and 17 over 4 inches. 178 of the smaller specimens were measured accurately with the following results:
The smallest had just absorbed the yolk-sac, and in many there was yet a small quantity of yolk remaining. Those 1.6 inches long had not been feeding over three weeks, as other observations show that they grow at the rate of .3 inch a month. That we found young salmon that had barely absorbed the yolk-sac indicates that they begin their migration as soon as they are able to swim. That they remained of the same size from October 8 to November 30 indicates that practically all start down stream at the same age. If any of the same age as those taken October 8 had been taken during the last days of observation, the fact would have been indicated by their increased size. In 53 days they would have grown over half an inch, and would have been at least two inches long. There were no intermediate sizes
between 1.8 inches and 4.0 inches. The larger individuals had remained over from the previous year. The trap was so set that they could not have avoided it during the day. The fact that none was taken during the day indicates that they travel more or altogether at night. The last record shows that 24 were taken from 8 to 9 a. m., which is the only day record. This is accounted for by high and muddy water, caused by a heavy rain during the previous night. This and other observations indicate that high and muddy water, especially the latter, is an incentive to day travel.
The two months' observations with the trap at Battle Creek were deemed of such importance that two observation stations were established on the river the first of January, 1899, and equipped with similar traps. The upper station was established at Balls Ferry, some three miles above the mouth of Battle Creek; the other at Walnut Grove on the lower river. The diagrams indicate the number of salmon fry taken in the traps at the two stations:
The diagrams indicate the passage of the fry from the fall run of adults; those from the summer run passed the upper station earlier, some of them being noted at Battle Creek during October and November. Observations in 1900 show that they may begin migrating in September.
The work at Balls Ferry indicates that the greater part of the young salmon passed that portion of the river between the middle of January and the middle of March. Practically all had passed by March 20. Numerous measurements made at frequent intervals show that the daily average size varied but one tenth of an inch during the three and a half months. The average of all measurements is 1.53 inches. This confirms the inference from the Battle Creek work that practically all the fry begin their migration as soon as they are able to swim, otherwise the later ones would have been larger.
Both diagrams show that a big run of fry is not necessarily coincident with a rise in the river. From the Walnut Grove diagrams we note. From the middle of January to the middle of May there were salmon fry in various numbers passing Walnut Grove. The height of the migration was from March 4 to about the 24th, lasting about 20 days. Practically all had passed by April 22. The size of those taken during January was 1.6 inches, during February 1.8 inches, during March 1.7 inches. From March 30 to May 7 they increased from 1.7 inches to 3.0 inches. The field notes show that as many fry were taken during the day as at night. The water of this portion of the river is muddy.
Comparing the two diagrams we note that a large run of fry passed Balls Ferry February 2, and that 34 days later, March 8, there was a large run of fry passing Walnut Grove. Again, a large run passed Balls Ferry February 14, and 34 days later, March 20, there was a large run passing Walnut Grove. The later Balls Ferry runs were caught by the high water the last of March, and their passage was not noticed at Walnut Grove. The high water probably carried them down faster, and they may have passed while the trap was out of the water, March 23 to 26.
The average size of the fry passing Balls Ferry January 30 was 1.51 inches, and of those passing Walnut Grove 34 days later 1.79 inches, an increase of .28 of an inch. Those passing Balls Ferry February 18 were 1.48 inches long, and 34 days later at Walnut Grove 1.77 inches, an increase of .29 inch. We have learned from other sources that the fry increase in length about .30 inch a month, and .28 and .29 inch in 34 days is not far from that rate.
It seems, therefore, both from the size and number taken at the two stations, that fry are about 34 days in passing from the upper to the lower station. The distance between the two stations is about 350 miles, as the river winds, and their average daily progress is therefore about 10 miles a day.
An object floating with the current would make the distance between the stations in less than 10 days, and if fry traveled only at night and merely kept with the current, they would be only 18 or 20 days on the way. It is probable that fry in migrating drift down stream tail first, keeping the head up stream for ease in breathing, as well as for convenience
in catching food. In this way they would drift more slowly than the current. I have seen fry at Battle Creek fishery traveling with the current, and always with the head up stream, unless frightened. The later and larger specimens found at Walnut Grove had simply been longer on the road. The larger they became the more slowly they drifted. Without doubt there are a few passing down the river all summer, though we have been unable to find any after June.
Much seining was done along the shores of Suisun, San Pablo and San Francisco bays, and in Tomales Bay, to determine something of the fry in brackish and salt water. Very few specimens were taken, probably not over 50. The smallest specimens taken in San Pablo Bay were 2.4 inches long, which indicates an age of about 414 months, and a period of three months for the migration.
A net stretched across the mouth of Olema Creek, emptying into Tomales Bay, caught salmon fry coming back into the stream at flood tide. This indicates that they regularly travel back and forth with the tides. We should expect this, as it is hardly probable that they know any directions, except with or against the current.
Many experiments were performed at the Hopkins Seaside Laboratory for the purpose of determining the effect of sea-water upon young salmon of various ages. 25,000 salmon were hatched in the laboratory, and at various ages a few were placed in separate tanks and subjected to various mixtures of fresh sea-water. Without going into details of the experiments, the following are the results obtained:
The young of any age can bear with impunity a density of 25 per cent, sea-water; that is, 1 part sea-water to 3 parts fresh water. Not until forty days old, at the time of the complete absorption of the yolk-sac, could they withstand a density of 50 per cent, sea-water. At the age of 50 days 75 per cent, sea-water could be endured. Pure sea-water could be endured at the age of 60 days, though there was a slight loss. It is doubtful whether they can enter salt water with complete impunity until 3 or 4 months old. The loss was much less when the density alternated from low to high and back again, simulating the change of density in the estuaries with the change of tides. In all cases the young salmon died when changed directly from fresh to salt water, or when the density was rapidly increased until that of seawater was reached.
To summarize the notes on migration: The fry begin their down stream migration as soon as they are able to swim. In clear water they travel more at night; in muddy water, as much or more during the day. Much of the time they float down stream tail first. In the larger streams they travel more or less in schools. Their regular migration is not dependent upon the height of the water, but upon age. Their rate of progress is about 10 miles a day, and they are about six weeks in reaching brackish water. They are probably four or five months old when they reach the ocean.
In the upper portion of Sacramento River and its tributaries there remains after the winter and spring migration a large number of young salmon. In the vicinity of Sims in 1898 we found 700 to 1,000 in the various pools. They were common in McCloud River in September, and in Fall River in August.
These summer residents, as they may be called, are confined to the headwaters, the clear streams with rocky bottoms. They do not stay much of the time in the swift current or riffles, but remain in the more quiet pools, where they feed on insects and take the angler's fly the same as trout. Considerable effort was made to learn as much as possible concerning them, and Sacramento River near Sims and its tributary, Hazel Creek, were visited each month from May to December with that end in view.
During July and August, all specimens taken were marked by cutting off the adipose fin, by which means we were able to make estimates of the number in the pools, their rate of migration, and their rate of growth. The number estimates were made thus: After having marked a few and released them in the pool, the following proportion was formed with the data from each seine-haul: the number of marked fry taken is to the number of marked fry in the pool, as the total number taken is to the total number in the pool. In August, when we could distinguish those just marked from those marked in July, we were able to make estimates of the number of July-marked fry in a pool and, knowing the number released there in July, to compute the rate of migration for the month.
The following table gives the result of the work in one pool, and illustrates the data used in making the number estimates:
|Date.||No. caught.|| No. previously
| No. marked
| Estimated No. |
|Aug. 16 P. M.||66||1||1|
|Aug. 17 A. M.||83||312||38||682|
|Aug. 17 P. M.||19||576||9||1216|
|Total different individuals,||Average of|
|24 hrs., 586||estimates,||1022|
The average estimates for two other smaller pools are 685 and 861. From these estimates it seems probable that there were about 10,000 young salmon to the mile in the upper Sacramento during the summer of 1898, or less than a million in all the headwaters, which is a very small percentage of the year's product for the river. At Battle Creek Hatchery alone that year, 1897-8, there were nearly 50,000,000 eggs taken. There is little migration of the young salmon during the summer months. In August we could account for 94 per cent. of the fry we had marked in one of the pools in July; 76 per cent. in the pool in which they were released, and 18 per cent. in the pool below.
The growth of the larger fry in fresh water is slow. The average size of the marked fishes in August was 3.91 inches, in September 3.86 inches, in October 4.20 inches, in November 4.20 inches, being a total gain of but.29 inch in three months. The earlier rate was.3 inch per month. Another effect of this summer residence in fresh water is the maturing of the male genital organs. A large proportion of the males that remain in fresh water until they reach a length of four inches becomes sexually mature. Its significance is not understood, and its effect upon the future growth of the fish is not known. The number of parasites found in the stomachs of young salmon living in fresh water increases with their size and with the time of year, there being more in the fall than earlier. Their food in fresh water at all ages, seasons, and places is insects, about two thirds being aquatic larvae, and the other third adult insects.
As stated above, young salmon reach the ocean when four or five months old. Concerning their habits from that time until they return to fresh water at maturity, we know very little. They are occasionally taken in the ocean near San Francisco, but so rarely as to indicate very little concerning their habits. They are abundant in Monterey Bay during the spring and early summer, but their appearance there is only the first step in their migration up the rivers to spawn.
The length of their stay in the ocean has been determined with considerable certainty by a series of observations carried on recently with the Columbia River salmon, which is the same species as the Sacramento salmon. In May, 1896, 5,000 young salmon 2.5 inches long were marked by cutting off the adipose fin, and were released in the Clackamas Biver, a tributary of the Columbia. The eggs from which they were hatched were spawned in September, 1895. During the summer of 1898, a little over two years after the marking, and a little less than three years after the spawning of the eggs, 375 of the marked salmon were taken in the Columbia, and five were taken in the Sacramento River in California. A few more were taken both in the Columbia and in the Sacramento in 1899, and again in 1900. The size of those taken in 1898 varied from 10 pounds to 57 pounds.
Besides indicating the age of the spawning salmon, this experiment shows that most salmon return to the river through which they reached the ocean. They do not do this because it is the stream in which the parent fishes spawned—the 5,000 mentioned in the experiment were from eggs taken in California—but because their ocean feeding grounds are not far off-shore, and in their two years' residence in the ocean they have not wandered far from the point at which they entered it. When it comes time to return to fresh water, their native stream is the first to attract attention.
Adult salmon may be found in the Sacramento River at any time of year. There are, however, two distinct runs, the earlier of which passes up the river during April, May and June, and the later during August and September. The former is known as the spring run and the latter as the fall run. The salmon of the spring run ascend the river to the headwaters, such as the upper Sacramento, McCloud River and Hat Creek, and some of the earlier ones even pass Pit River Falls and enter Fall River. The salmon of this run spawn mainly in August. The fall salmon do not ascend the river so far as those of the spring run, but turn aside into the lower tributaries, or spawn in the main river. They reach their spawning grounds during the latter half of October, November, and the first half of December. The main river is very low at this time of year and only a small portion enter the tributaries.
Details of Migration.
When the salmon enter the bay from the ocean, they come in against the ebb-tide. They stem the current till the tide changes, and then run out against the flood-tide, losing much of the distance gained during the ebb. That they do not lose altogether as much as they gain may be understood from the following explanation:
The tide runs up the bay and river as a broad low wave, on the upper side of which is flood-tide, and on the lower side ebb-tide. This wave is about three hours going from San Francisco to Benicia; it reaches Collinsville in about four hours, and Rio Vista in four and a half hours. When the crest of a wave is at Isleton, its trough is about at the Golden Gate. The farther the tide extends up stream, the smaller the wave, the shorter the flood, and (as the flood and ebb must together equal' twelve hours) the longer the ebb. The following diagram will illustrate the movements of a salmon in passing through the bays: a, b and c represent the tide-wave at successive points as it passes up the bay. <-< indicates ebb-tide, and >-> flood-tide. Suppose that a salmon enters the Golden Gate, G G, at the beginning of ebb-tide, which would be the most favorable time. His position on the wave will be at S. If he is able to travel up the bay as fast as the wave he will keep his position near the crest, that is at S. But he can hardly do that, especially as Map of Sacramento River and Outlet, showing stations, by names or numbers, where salmon inquiries have been made. the current would be very slight, and in the broad bay hardly strong enough for his guidance. Let us suppose that by the time he reaches Benicia, B, he has fallen behind the wave until he has the position at S'. It is then slack low water, and he can make no headway. Soon the next wave reaches him and he is in flood-tide. He will therefore swim back against the current. As the wave is going up the bay and he is going down, he soon gets past the crest and finds himself in the ebb-tide at S". He then turns and stems the ebb-tide, and as the wave is going in the same direction he is, he goes much beyond Benicia, B, before he again falls back to slack low water at S'", and gets into the flood of the next tide-wave.
By taking the statistics of the daily catch of salmon at various points, we have been able to trace the progress of a school up stream, and find that it requires four days to pass from Vallejo to Sacramento. We have been unable to determine the length of time required to reach Vallejo after leaving the ocean.
The spring run passes up the river quite rapidly, reaching their spawning grounds in the McCloud River in about six weeks after entering the river at Collinsville. The fall run moves more slowly. The flood and ebb tides are more nearly equal, owing to the small amount of water coming from the rivers, which makes a longer passage through the bay. After reaching the shoals in the middle portion of the river they move slowly, having already found pretty good spawning grounds. They are about two months reaching their spawning grounds (between stations 17 and 8 of the accompanying map), which are but little more than half as far up stream as those of the spring run.
In September, 1901, over a hundred salmon were weighed and branded with serial numbers and released in the river at Rio Vista. Three of these were taken later upon their arrival at the spawning grounds. The following is a tabular statement of the data concerning them, the loss in weight being due to migration alone. No. 34 was 8 days in spawning, extruded all but 20 ova, and lost thereby 21 per cent, more of the Rio Vista weight.
The salmon of the spring run reach their spawning grounds from two to six weeks, or even longer, before spawning, which time they spend lying quietly in the pools. The fall salmon are more nearly ripe when they arrive at the spawning grounds, many of them ceasing to ascend only when ready to spawn. Head of a Salmon with the Opercle cut away, showing the gills attacked by fungus and Parasitic Copepods. Yet alive when taken.
Spawning and Death.
Salmon are in fairly good condition when they begin spawning, notwithstanding the fact that they have been without food for several weeks or even months, and have traveled several hundred miles in the meantime. The male has changed his appearance. His snout has grown longer and much hooked; large, conical, hooked teeth have appeared in his jaws; his body has grown thinner and apparently deeper, though the latter point has not been determined yet by measurements; his skin has thickened and embedded the scales. The larger males become somewhat reddish in color as spawning time approaches. The female has changed less. The abdomen is somewhat distended from the ripening ova, the skin thickened as in the male, and the color has become more or less of a dull olive.
As spawning progresses, both sexes rapidly become emaciated. Their fins become frayed and worn, especially the caudal fin of the female, which is frequently entirely worn off. Fungus attacks the skin in various places, especially the top of the head and the fins, and it frequently destroys one or both eyes. The gills are nearly always
|Alimentary Canal of a Salmon taken in Monterey Bay.||Alimentary Canal of a Salmon after Spawning, taken at Battle Creek Fishery|
attacked by fungus and parasitic Copepods, and half or even three fourths of the gill filaments are sometimes thus destroyed. The skin is worn off in many places, such as the side of the tail, the projecting edges of the jaws and fins, and the snout of the male. They have eaten nothing since leaving the ocean, and the alimentary canal has
long since shriveled up until the stomach is scarcely one tenth its normal size; a catarrhal desquamation takes place all along the digestive tract, but especially posterior to the stomach, which portion is frequently filled with tapeworms. At least two other kinds of parasitic worms are frequently found in the abdominal cavity. Both sexes lose from 30 to 40 per cent, of their weight.
Such is the condition of the salmon at the time of death. It is not the spawning itself that produces this effect, but the continuation of the spawning efforts after all ova have been extruded and fertilized. Usually there is considerable vitality in both sexes at the time of the complete extrusion and fertilization of the ova, and they have enough energy left, in many cases, to carry them back to the ocean. But they exhibit no inclination to go. Instead, they continue on the spawning beds, persevering in their efforts to spawn and fertilize more eggs. When one fish of a pair dies, the other seeks or is sought by a new mate. Probably the female never notices the death of the male, as there are nearly always several males fighting around each 'nest.' If there happens to be no salmon in the vicinity when one of a pair dies, the
survivor continues his or her efforts alone, futile though they are.
This extreme devotion to a purpose is almost without parallel. There is no instinct whatever to return to salt water. It is even doubtful whether they would revive if taken back. The dog salmon and bump-back salmon spawn in small streams that empty directly into the ocean, and they are found dying and dead in brackish water. The fact that all salmon, of the genus Oncorhynchus, die immediately after spawning once cannot be questioned.
- The description here given of the spawning salmon should not prejudice any one against the fish as found in the markets, either fresh or canned. Such are taken either in salt water or within a few days after leaving it, and are in prime condition. Indeed, they must necessarily be, in order to make the long journey up the river and live several months without food. Chemical analyses have shown that the Sacramento salmon is the most nutritious of fishes.