S44 
FOREST AND STREAM, 
some variation of which the word "Red" is always a part; 
the silver salmon is usually "Coho," or "Medium Red;" 
the humpback is called "Pink Salmon," and the dog sal- 
mon is "Chum." Not infrequently the inferior grades 
are put up under fictitious labels, in many cases merely 
deceptive, in others actually false. 
While the pack of Salmon in Alaska this year exceeds 
2,000,000 cases, valued at more than $8,000,000, it is 
somewhat less than that of 1902. This large pack was 
secured only through unusual energy and persistence on 
the part of the canners, and it is of vital importance that 
the Government take steps at once to preserve these im- 
portant fisheries. Barton Warken Evermann. 
Sawdust and Fish. 
BY A. P. KNIGHT, M.A., M.D., PROFESSOR OF ANIMAL BIOLOGYj 
queen's university, KINGSTON. 
(The followmg report on the effects of sawdust on fish relates 
to experiments undertaken by Professor Knight, at the su^^estion 
•of Professof Prince, Fish Commissioner for the Dominion of 
Canada.) 
The source of the poison given off by sawdust is 
undoubtedly to be found in the contents of the wood 
ceils. Sugar, starch, oil, resin, gum, jelly, alkaloids, 
and acids are all examples of material stored in dif- 
ferent parts of plants. 
In the older parts of trees the protoplasm and sap 
disappear completely from the cells, and they may then 
contain nothing but the stored material. In the pine 
family there is stored in the wood and bark cells an 
abundance of crude turpentine and resin. The Norway 
spruce of Europe furnishes in this way turpentine and 
Burgundy pitch. The yellow pine of the Southern 
United States yields spirits of turpentine by distilla- 
tion of the crude turpentine which runs away from the 
tree by cutting into it. The residue after the distilla- 
tion is resin. 
Now, the poisonous material in sawdust must be 
either the cell wall or the stored material. It cannot 
be the cell wall, for this is just the wood fibre or ma- 
terial used in making paper, and pure paper is certainly 
not harmful to fish life. The poison can scarcely be 
anything elese than the turpentine and other sub- 
stances stored in the cells. 
Different trees, such as tamarack, pine, cedar, spruce, 
etc., generate and store different kinds of reserve ma- 
terial. When a log from one of these trees is cut into 
boards, the sawdust gives off proportionately much 
more poisonous matter than the slabs, edgings and 
bark. The reason of this is easily understood. As 
each cell or vessel is microscopic, and contains only 
a very small quantity of poison, and as the cell wall 
must be broken open in order to let out the contents, 
St follows that the greater the number of cells that are 
opened, the greater will be the quantity of turpentine, 
resin, etc., poured out. Hence, a saw log converted 
into sawdust, or ground into shreds, as in a pulp mill, 
gives out the maximum of poison; whereas a similar 
iog sawn into boards, edgings and slabs, will give out 
a much less quantity. The minimum will be given out 
by a saw log floating in the water. 
There are other industries in Canada which in pre- 
paring their products for market grind up plants and 
trees, and thus let out their ceil contents. One of these 
is the pulp industry — likely to become very extensive 
in the near future. Two processes are in vogue in 
this industry. In one, the logs are macerated with 
chemicals, the mills being known as sulphite mills. In 
the other process, the logs are ground into shreds in 
what are known as mechanical mills. Both processes 
liberate the greatest possible quantity of stored ma- 
terial from the wood cells, and if this material is 
equalljf poisonous with that liberated from sawdust, 
then the waste water discharged from a pulp mill 
should be much more poisonous than from a sawmill. 
The St. Andrew's experiments determined the percent- 
age of poison from a sulphite mill which is fatal to 
fish life, but, so far as I know, the percentage of poison 
from a mechanical mill has never been determined. A 
provisional conclusion, however, may be based upon 
some of the experiments to be described later in this 
paper. 
The experiments hitherto described would seem to 
indicate that some considerable time was required for 
the water to dissolve out the poisonous extracts from 
white pine sawdust, but such is certainly not the case. 
Millmen and anglers alike testify that many kinds of 
fish are taken by hook and line at mill-ends, no matter 
how excessive the sawdust may be. The sawdust does 
not kill the fish so long as there is a rapid and 
abundant flow of water. Why do fish thus congregate 
at mill-ends? To answer this question we must re- 
member two things: first, rapidly ruiming water is 
better aerated than sluggish water; and secondly, some 
fish, such as trout and salmon, ascend streams until 
they reach suitable spawning grounds, or are stopped 
in their ascent by high falls or milldams. In ascending 
a river these fish are but obeying a law of their nature; 
in- congregating at mill-ends they are equally obeying 
a law of their nature, and are instinctively seeking 
water which furnishes their blood with a plentiful sup- 
ply of oxygen. This instinct is well illustrated in one 
of his experiments. The experiment was repeated a 
number of times, and in every instance the fish dis- 
covered where the fresh water came in. In one in- 
stance, in order to get close to the incoming water, 
a minnow stood on its head for fifteen minutes with 
more than half of its body buried beneath the sawdust. 
It was thus acting under the impulse of two funda- 
mental instincts, viz., the instinct to avoid poisoned 
water on the one hand, and to seek fresh water on the 
other._ The experiment seems to throw light upon the 
experience of anglers who have found that trout desert 
the main stream when saw mills are running, and be- 
take th emselves to the unpolluted branch streams 
lower down. 
Reference has already been made to the fact that 
black bass fry, minnows and perch, when placed in an 
aquarium, invariably avoided the poisonous sawdust 
water at the bottom. Having sunk into it once or 
twice, it was found almost impossible to drive them 
into it again. Here was a conflict between two funda- 
jijent^I iristincts. the one haijd was the natural in- 
stinct to hide in deep water; on the other hattd, the 
equally natural instinct to avoid the poisonous solution 
at the bottom. Which instinct would the fish obey if 
compelled to make a choicfe? 
The following experimfertt Was designed for the pur- 
pose of seeing which instinct was the more powerful, 
and for the further purpose of imitating what might 
possibly occur in a stagnant pool along the course of 
a sawdust polluted stream. 
A glass aquarium I2in. x Sin. x 6in, was placed in a 
much larger vessel and a mixture of ice and salt packed 
in the latter so as to surround the aquarium. The 
aquarium was then half filled with white pine extract 
which had been forming for three weeks, arid which 
killed adult fish in from one to three nlinuties. Aftfer 
the extract had been cooled dowri to 8° c,, tap water 
at the temperature of 13° Was slowly admitted to the 
aquarium so as not to disturb the underlying pdisorldus 
water. The tap water, being warmer, floated clear and 
transparent on the dark purplish extract below. The 
clear water entered and left the aquarium at the rate 
of igo c.c. per minute. 
At first two minnows were placed in the aqiiarium. 
They at once dove to the bottom, encountered the 
poisonous water, immediately came up again, repeated 
the operation a few times, and finally remained swim- 
ming about in the clear water. Three black bass fry, 
liberated one after the other, went to the bottom and 
never came up — suffocated and poisoned in the dai'k 
stagnant water at the bottom. Of two other minnows 
dropped into the aquarium, one large one never came 
to the surface; the other joined its fellows in the clear 
water above. All three soon found the end at which 
the fresh water was entering and remained there facing 
the stream. 
This experiment shows what might possibly happen 
in pools partially filled with sawdust. Wood extracts 
would form, and being cooler and heavier than the 
clear water, would lie at the bottom of the pool. Of 
course, fish already in the pool Avould be driven away, 
but those coming up or down stream through shallow 
stretches, and trying to hide in the deeper waters of 
the pool, might be suffocated or poisoned. 
One objection frequently urged against the practice 
of throwing sawdust into streams and rivers is that the 
decaying sawdust imparts such a disagreeable odor to 
the Avater that sensitive fish are driven away to other 
waters not so polluted. It seemed to me, therefore, 
that some progress might be made towards a definite 
conclusion in this matter, if sawdust were allowed to 
stand for several weeks in an aquarium and tested from 
time to time as to the changes going on in it, and the 
influence of these upon fish. 
With this end in view about tooo grams of white pine 
sawdust were placed in an aquarium three feet four 
inches long, fifteen inches wide, and filled up to sixteen 
and a half inches deep with fresh water. This was done 
June 24. No water was allowed to enter or leave the 
vessel. No direct sunlight fell upon it: 
The usual results followed, viz., a well defined layer 
of pale, yellow water one and three-quarter inches deep 
formed in a few hours and lay at the bottom. On top 
of this was the perfectly clear layer about fifteen inches 
deep. 
After soaking for two days, bubbles of gas began to 
rise to the surface of the water, but no attempt was 
made to analyze it. The bottom yellowish layer had 
become so dense that no object could be seen across 
it — a thickness of fifteen inches. Its upper surface was 
sharply marked off from the overlying transparent 
water by a thin grayish layer. Microscopic examina- 
tion of this layer showed it to be swarming with bac- 
teria. 
At the end of a week, only about an inch at the bot- 
tom had retained the original yellow color; the next 
inch had changed to a yellowish brown; then came a 
grayish layer about one-sixteenth of an inch thick; 
above this, what had at first been fourteen inches of 
perfectly pure water had turned to a dark gray, though 
still quite transparent. Black bass fry placed in the 
aquarium at this time at first sank to the bottom, but 
after meeting the poisonous extract once or twice 
could not subsequently be driven into it. On the con- 
trary they swam along the top with their nose just 
touching the surface of the water, and behaved as if 
suffering from lack of air. They lived only about two 
hours. 
Four days after this, black bass fry placed in the 
upper fourteen inches lived only about one hour. They 
also swam along the surface and appeared to be gasp- 
ing for air. That they were suffocating in both cases 
was proved by the fact that when fry were placed in a 
wash bottle of this water with air bubbling through it, 
they lived on for twenty-four hours, and were then ap- 
parently well and exceedingly active. On being trans- 
ferred from the wash bottle to the aquarium the ani- 
mals at first plunged downward to the bottom, paused 
there a moment, but soon came toward the surface 
breathing very rapidly. Evidently they were suffering 
from lack of oxygen. They swim along the top with 
noses upward and body inclined at an angle of about 
thirty degrees with the surface. Gradually they tire; 
sink toward the bottom; rise again; swim convuls- 
ively toward the surface; jump clear out of the water 
with gaping mouth; become exhausted by their con- 
vulsive efforts and finally sink to rise no more. Of all 
the fish killed in this extract not one ever rose to the 
surface after death. 
It would be difficult to say whether this experiment 
throws any light upon a point much discussed in the 
literature of sawdust. The point is this: if sawdust 
kills fish, why are they not found dead in considerable 
numbers along the course of the stream? In my ex- 
periments the dead bodies of the fish never rose out of 
the poisonous liquid. 
However, the connection between a few links in the 
chain of animal life was apparent enough, viz., wood 
extracts supported bacteria, bacteria supported mos- 
quito larvse, and these again (after aeration of the 
water such as would occur in running water) supported 
fish life. These observations dispose to some extent 
of the oft repeated charge against sawdust that it- de- 
stroys the food of young or newly hatched fish. When 
roionows reljshgd mosquitg larvae as foqd, and I fff' 
quentiy saW tliehl eatinjg the larvjE, it requires rio great 
stretch of the scientific irrtagination to understand how 
fish fry of different kinds, such as trout and salmon, 
might subsist upon the larvae of mosquitoes and other 
aqiiatic insects, these latter in turn subsisting upon bac- 
teria, and the bacteria subsisting upon the organic mat- 
ter derived from the decaying vegetation of the forest. 
Another thought comes up in connection with the 
presence of organic matter in streams and rivers. 1?lie 
organic matter which passed into a river when Canada 
was covered with forest riiuSt have beeh c^uite different 
in character firom that which this saMe streani receives 
today from the vegetation of the farms along its valley. 
The surface drainage from a forest must differ in kiil4 
frbni the Sul-fage drainage of a farm, arid the bacterial 
life in each must differ also. Moreover, tbe waters of 
our smaller streams were, years ago, shaded by trees, 
and the varieties of their bacterial life must thus have 
been quite different from the bacterial life in sunlit 
streams of today. Consequently, it may fairly be 
argued that the insect life, in and along the streams of 
an agricultural district, differs both in kind and number 
from what characterized these same streams 100 or 200 
years ago. And if larval and adult insect life has 
dwindled or disappeared, so must the fish life which 
subsisted upon it. 
The Anglo-Saxon has always been a disturbing fac- 
tor in the balance of life. Forests, game and fish all 
disappear with his arrival. To get good fishing or 
good hunting nowadays one must travel back to un- 
settled districts. No one expects game to be plentiful 
along the shores of Lake Ontario, but many people are 
amazed that fish are not abundant in it. They still 
hug the pleasing delusion that if brooks have been 
overfished, the fish hatchery can re-stock them. But ' 
with the disappearance of our forests it is exceedingly 
doubtful whether we can ever again, by all the help of 
hatchery, overseers and fish commissioners, re-people 
the streams which have been depleted by man through 
over-fishing and deforestation. He has upset the bal- 
ance of life; it can only be fully restored by a return to 
primitive conditions. When game, therefore, becomes 
plentiful on the streets of Ottawa city, fish will be 
equally abundant below the saw mills of the Chaudiere 
Falls. 
Such, at least, is the conclusion to which my experi- 
ments point, notwithstanding the indisputably poison- 
ous effects of strong solutions from sawdust near the 
source of pollution. As I have already pointed out the 
question of whether any particular stream is suffi- 
ciently polluted with sawdust to kill fish life is simply 
the question of determining whether enough sawdust is 
passed into the stream to poison its waters. The for- 
estry engineer will soon be trained to determine the 
strength of sawdust solutions, and will then be able to 
settle this question of pollution beyond the possibility 
of doubt. 
At present, however, a final judgment cannot be pro- 
noimced upon the poisonous effects of sawdust. These 
effects must be studied near the mills and along the 
sawdust beds of our rivers. A three weeks' study of 
the Bonnechere river, a tributary of the Ottawa much 
polluted with mil! rubbish, led me to modify very con- 
siderably the conclusions which I had based upon my 
laboratory experiments. I visited the mill, fully ex- 
pectmg that not one fish could survive in such sur- 
rouwdings. But pike were abundant for miles below 
the mill, and fish (chub) could be caught any day along 
the side of the submerged driftwood. Stranger still, 
the fish so caught lived for three hours in a pailful of 
sawdust water drawn from the very center of a sawdust 
bed. A few brook trout had been caught earlier in the 
season just below the mill when it was running. At 
the date of my visit, August 20, 1902, the mill had 
been closed for seven weeks and no sawdust Was then 1 
passing into the river. 
Of course, these figures are mere approximations, 
but they point unmistakably to the conclusion that the 
sawdust poured into the Bonnechere river is not de- 
stroying its fish hfe. Moreover, in Golden Lake, an 
expansion of this same river, and ten miles above any 
saw mill, lake trout used to be very abundant. Every 
October large numbers were caught in nets along their 
spawning beds. Now these spawning grounds are re- ' 
ported to be deserted by the fish, and certainly sawdust 
cannot be blamed for their disappearance. Higher up 
the river, in Round Lake, the October fishing is still 
good, solely because there are fewer settlers and less 
fishing. 
1. Strong sawdust solutions, such as occur at the 
bottom of an aquarium, poison adult fish and fish fry, " 
through the agency of compounds dissolved out of the ; 
wood cells. 
2. The overlying water in such an aquarium does not 
at first kill fish. After about a week it does kill, but 
solely through suffocation, the dissolved oxygen hav- 
ing all been used up. 
3. Bacteria multiply enormously throughout all parts 
of such an aquarium, and through oxidation change the 
poisonous extracts to harmless compounds. Mosquito 
larva; live on the bacteria. No doubt, in natural pools, 
other aquatic insect larvae live on bacteria also. 
4. Subsequent aeration and sedimentation of sawdust 
water purify it, so that fish can live in it without injury. 
5. Since adult fish and black bass fry both refused to 
be driven into pine extracts in the bottom of an aqua- 
rium after they had experienced its poisonous effects, 
we may infer that fish would desert a river much pol- 
luted with sawdust, going down stream and into tribu- 
taries to escape from the disagreeable influence of saw- 
dust extracts. 
6. No stream can be pironounced off hand as poison- 
ed by sawdust. Each stream must be studied by itself 
and the varying conditions must be thoroughly under- 
stood before a judgment can be pronounced. The chief 
thmgs to be considered are (i) the quantity of sawdust, 
and (2) the volume of water into which the sawdust is 
discharged. Subordinate conditions are the rapidity or 
shiggishness of the stream, the amount of sunlight or 
shade, and the character of the water, whether from 
agricultural lands or from primitive forests. 
7. Further observations and studies along sawdust 
polluted streams and rivers of Canada are urgently 
l]CCt}?-4 before more definite cotj^jlqsioris c^q be reached, 
