596 
WARURE 
[APRIL 22, 1897 
it with an increased range in the passage between Gaspé and 
Anticosti, and from there pursues its way with ever-increasing 
height up the St. Lawrence to (Quebec. The progress of this 
tidal wave has been traced to the existence of a deep channel 
which crosses the whole extent of the Gulf from Cabot Strait to 
the passage between Gaspé and Anticosti, and thence up the 
St. Lawrence nearly to Saquenay. This channel extends a dis- 
tance of 500 miles, with an average width of 35 miles and a 
continuous depth of 150 fathoms. 
For the purpose of ascertaining the nature and velocities of 
the currents, the steamer used was moored with a wire rope 
hawser, provided with an accumulator, to prevent sudden jerks 
and strains. This accumulator consisted ofa series of sixty rubber 
discs, five inches diameter, making a total length of twelve feet, 
which was reduced to eight feet eight inches under the greatest 
compression. Two kinds of current meters were used: one 
having small buckets revolving horizontally, on the same prin- 
ciple as an anemometer ; and the other of a fan, similar to a 
screw propeller, revolving in a vertical plane. The former was 
found to be best for sea work, as it was least affected by the 
vertical motion of the vessel due to waves. The latter was found 
to be liable to head up or down as the vessel rolled, and so give 
an exaggerated record. Both kinds were worked by electricity. 
The surface currents were taken at a uniform depth of eighteen 
feet, which was well clear of the keel of the steamer. The 
meter was allowed to run for half an hour at this depth, then 
lowered to the desired depth for an hour, and then again run for 
half an hour at the eighteen feet. At a depth of ten fathoms 
the Gaspé current was sometimes found to be stronger than at 
the surface, but usually the velocity decreased regularly with the 
depth. At twenty fathoms it was only 50 per cent. of the 
surface velocity; and at thirty fathoms, 20 per cent. The 
greatest velocity was 2°81 knots. The current fluctuated with 
the rise and fall of the tide, decreasing during the rise, and in- 
creasing during the fall. The constant outward current from 
the Gulf was found to have a width of fourteen miles, and a 
depth of forty fathoms near the Gaspé coast, with a surface 
velocity of from 1°10 to 2°81 knots. The temperature in July was 
found to range from 53° at the surface to 33° at thirty fathoms ; 
and 32° at fifty fathoms. In the Gaspé region and in Cabot 
Strait the coldest water formsa layer between the depth of thirty 
and fifty fathoms, and, while the surface water rises in tempera- 
ture during the season, no appreciable variation was found from 
June to September at a depth of fifty fathoms. 
The density of the water was ascertained as affording an indi- 
cation of the quantity of fresh water coming down the St. Law- | 
rence. In the Strait of Belle Isle and Cabot Strait the density 
of the surface water ranges from 1°0233 
in the Atlantic. On the western side of Cabot Strait, the out- 
flowing water, which occupies a width of about ten miles, has a 
density of 1'0220 to 1'0235 at the surface. In the Gaspé region 
the average density for a width of fourteen miles, and between 
the surface and ten fathoms, was 1°02195: and to a depth of 
forty fathoms, was 1°02368, The density of the water is dis- 
turbed by currents due to wind. Thus during three days, when 
the wind from the S.S.W. averaged twenty miles an hour, the 
density contours were displaced to the northward about nineteen 
miles at the surface, fifteen miles at ten fathoms, and nine miles 
at twenty fathoms. 
The completion of the survey is expected to occupy another 
three or four years. 
222 
AGRICULTURAL EXPERIMENTS IN PLOTS 
AND POTS. 
Na recent number of the Avrécau/tural Gazette of New South 
Wales (vol. vii. p. 663) there is an article by Mr. N. A. 
Cobb, written at the request of the Minister for Agriculture, 
upon the methods employed for experiments with crops and 
manures. It appears that field experiments are being carried 
out toa considerable extent by the farmers of the country, but that 
the results are to a large extent untrustworthy and misleading, 
owing to innumerable sources of error which the experimenters 
have failed to perceive and guard against. Science is thus 
brought into ill repute, doubt is thrown on established truths, 
and progress hindered. The evidence brought forward goes 
far to show that this is a true indictment. When, however, 
the author goes a step further, and speaks of field experi- 
to 1°0245, the same as | 
him. The sources of error which he mentions may all be 
avoided by judicious management, if only the experimenter will 
guard against them at the commencement of his work, and 
superintend his operations with proper care. 
Inequalities of soil are one of the worst evils in field experi- 
ments ; the investigator frequently remains unconscious of them, 
the difference in the results being credited to the effect of the 
manures, &c. It is vexy rare for proper precautions to be taken 
against this evil, for the simple reason that these precautions 
imply delay, and the experimenter is generally in a hurry to 
obtain results. If, for instance, the comparative effect of dif- 
ferent manures on barley is to be ascertained, or the compara- 
tive yield of different varieties of seed, the only basis for an 
accurate trial is to divide the field into the required plots, then 
sow the whole field with a uniform barley seed, without any 
manure, and weigh separately the produce of each plot. If the 
crops obtained are equal, within the unavoidable errors of ex- 
periment, the field is one suitable for the purpose of the experi- 
ment; if the crops are unequal, the field, or that portion of it in 
which the inequality occurs, is clearly unsuited for the purpose 
intended, It is not sufficient, as is often supposed, to inspect the 
field when under ordinary culture, and because of the apparent 
evenness of the crop, to pronounce it fit for use; for natural 
inequalities of soil may not appear in a well-manured field, 
although plainly manifested when the supply of manure ceases. 
The errors due to inequalities of the soil in one series of trials 
may, of course, be neutralised by making many series of trials, 
and substantial accuracy may be gained by simply regarding the 
mean results obtained ; but if a field is really unequal in fertility, 
no ordinary arrangement of duplicate plots will suffice to ensure 
an accurate result. If the same experiment is repeated through- 
out a wide district, as is often now done in County Council 
experiments, it may be quite misleading tc take the mean of all 
the results as expressing the truth for the whole district. We 
must not bring into the mean the results obtained in different 
soils and climates, unless, indeed, our aim is to procure general 
statistics which are of no value for any particular place. Basic 
slag and superphosphate will compare quite differently upon a 
clay and upon a chalky soil; nitrate of soda and sulphate of 
ammonia will compare differently on dry and wet soils. To 
take the mean of experiments made under such different con- 
ditions is simply to misinform every farmer in the district; yet 
public money is continually wasted in this way. 
Mr. Cobb points out that the effect of inequality in the soil 
may be obviated by substituting rows for square plots. This is 
true, and the point is well worthy of attention ; the suggestion 
is not, however, novel. Ina comparison of basic slag and super- 
phosphate for turnips, conducted by the writer at Rothamsted in 
1886, the slag and turnip seed were sown by drill on the top of 
two ridges down the whole length of the field, and on the return 
of the drill an equal number of ridges by the side of the first 
were left unsown. When the sowing of the slag was completed, 
the same drill sowed superphosphate and turnip seed in all the 
vacant spaces. There were thus throughout the field two rows 
of turnips with slag, side by side of two rows of turnips with 
superphosphate, the repetition occurring many times over. This 
plan was suggested by Sir John Lawes. This is, for many ex- 
periments, a good mode of work, but its use is practically limited 
to those crops and manures which can be sown by drill; 
unfortunately, drills are not satisfactory machines for evenly 
distributing given weights of manure over given areas. 
Mr. Cobb next passes to the pot system of experiment: he 
describes the work at the Darmstadt Experiment Station, with 
its 1000 pots, and suggests that work on this system should be 
commenced in Australia. : 
There is no doubt that for solving certain questions the pot 
system, when carried out with scrupulous accuracy, is far 
superior to any other. If we wish to know what is the com- 
parative value to any plant of various nitrogenous manures under 
the most favourable conditions of supply and use, we arrive at 
this fact only by pot experiments. The produce obtained per 
unit of nitrogen in the pot will not, however, necessarily be the 
produce obtained in the field ; and the relative value of different 
manures, as shown in the pot, will only by mere chance appear 
in the field, where, in fact, it will be found to vary every year. 
The essential difference between the two systems is due to the 
fact that the field results are largely influenced by the season, 
and especially by the amount of rain, and the quantity of water 
percolating through the soil; while the pot cultures are care- 
ments as almost essentially untrustworthy, we cannot agree with | fully protected from such vicissitudes. If, then, the farmer 
NO. 1434, VOL. 55 | 
