I will but add a few very general conclusions I have arrived at from the foregoing.

First, the elongated ball possesses, from its very shape, a tendency to pursue the direction of its axis. This tendency should be made available, so far as practicable, to sustaining the flight, and flattening the trajectory.

Second. In the modern improvements in the art of throwing projectiles from rifled fire-arms, a decided step has been made backwards, in losing that most essential element to range and accuracy, initial velocity. It is desirable, and, I should think, not impracticable to restore it.

The initial velocity of the old rifle ball (weighing but about 180 grains) was 1750 feet per second. To the modern projectile, weighing about 500 grains, is given but from 900 to 1000 feet initial velocity.

It would doubtless be difficult to give the high velocity to so heavy a ball, and if given, the recoil would be inadmissible. But why throw so heavy a ball?-and why adhere to such calibres? The long range of such balls is not due to their weight, but to their model and to the low ratio of their cross section to that weight.

Both these advantages can be attained with a light ball as well as with a heavy one.

An increase of weight over that of the old musket ball is, in itself, objectionable, by increasing the soldier's load.

I can see no reason why, to a ball of the weight of the old musket ball (340 grains), may not be given all the properties of the heavier ones, with the additional great advantage of a high initial velocity, approximating to that of the old rifle balls.

To accomplish this, of course, the calibre must be greatly reduced the ball elongated, and a comparatively large charge used (less, however, than that used in the old musket-viz. 110 grains).

This highly elongated ball, balanced (if I may use the term) by grooves around the after end, so arranged that the resultant of the air's resistance shall pass as nearly as practicable through the centre of inertia,* will fulfill likewise the first condition I

* According to my theory of the matter, these grooves, instead of acting like an arrow's feathers, to keep the axis coincident with the tangent to the trajectory, have just the contrary object, to keep it (if properly adjusted) parallel to that original direction, by so balancing the forces of resistance as to cause their resultant to pass through the centre of inertia. The theory of "arrows' feathers" supposes, on the contrary, that, by the grooves, this resultant is thrown behind that centre, and tends to tilt the axis; an operation which must inevitably, likewise, produce deviation. Prof. Magnus' observation on balls with low initial velocities (so as to be visible to the eye) confirm my position, though I attach little value to those observations.

Experience seems to show that this balancing of the forces of resistance may be produced without grooves (vide Wilcox, p. 160), for the Enfield ball and the Whitworth projectile seem to be without them, as likewise Lancaster's. I have seen no statements, however, as to the deviation of these projectiles.

have laid down. For the highly elongated ball possesses in a proportional degree the tendency to pursue the direction in which its axis points: and the causes of deflection of the axis' direction being eliminated by a proper balancing about the centre of inertia of the forces of resistance, the tendency is to pursue the original line of direction, in opposition to the downward curvature due to gravity, and thus to flatten the trajectory and increase the

range.

These conclusions are those of theory alone, but, if I mistake not, all the most recent advances in rifled arms have been in the direction which they indicate. The Swiss (Federal) rifle, one of the most perfect in Europe (see Lieut. Wilcox's "Rifles and Rifle Practice," p. 187) has a calibre of but 0:41 (inch), and its ball, 2 calibres in length, weighs but 257 grains, thus combining with the small calibre, the highly elongated form, and even then weighing less than our old spherical musket ball. Though its initial velocity is not given, yet as the greatest proportional charge of powder is used with it, doubtless it also receives the highest initial velocity,* of any of its class of projectiles now known in any service.

The hexagonal projectile of Whitworth is another instance in confirmation of the principles I advance.

Of small calibre and highly elongated,† he throws this projectile with such accuracy as to hit with certainty, at 500 yards, a disk not more than two inches in diameter; and "asserts that he will not rest satisfied till he has fired a ball from one of his guns into the barrel of another, at a distance of 500 yards." (Edinburgh Review, Ap. 1859.)

The "Armstrong" projectile is another characteristic instance. So decidedly is the elongation of the ball characteristic of the most recent and successful efforts in obtaining range and accuracy, that the English writer just quoted applies the term "bolts" to the Whitworth and Armstrong projectiles.

In fact, the two springs from which have risen the modern improvements in projectile weapons are 1st, the application of the rifled principle to all arms; 2d, the elongation of the projectile. Either one alone may produce, to a certain extent, the results desired; it is only by the best possible combination of the two that the best results can be educed. The increase of range is due

*Initial velocity is the very first element in procuring range and accuracy. The greater it is, other things being equal, the flatter will be the trajectory (one of the principal elements of accuracy), and the flatter the trajectory the less the lateral deviation laterally. The extreme curvature of trajectory (though it may have its advantages in some peculiar circumstances), is the great difficulty as to accuracy of fire, in modern rifled weapons.

A necessary evil with the heavy balls used; the pretence that it is, in itself, an advantage, is too absurd for controversy.

According to Wilcox the mean calibre is about 046, and its length 2 calibres.

almost entirely to the latter principle, and it is only by applying it to the utmost practicable extent (as in the case of the Whitworth and Armstrong projectiles), that the greatest range and most perfect accuracy can be obtained.

The calibres in use have been a positive bar to the successful use of this principle in small arms. Borrowed from the old smooth-bored weapons, the adherence to them has caused an unnecessary increase of weight,* and made a loss of initial veloc ity inevitable, with all the attendant evils of a highly curved trajectory, and large deviation.

"Notwithstanding" (says Wilcox) "the long time that has elapsed since the discovery of the rifle, its principle is not yet so well understood as to have led to the general adoption of any particular form of this arm as the best."

The above conclusion of the author of "Rifles and Rifle Practice," will justify me, I hope, in venturing to make the foregoing suggestions on the subject.

New York, Dec. 15, 1859.

ART. XVIII.-Gulf Stream Explorations-Third Memoir. Distribution of Temperature in the Water of the Florida Channel and Straits; by A. D. BACHE, Sup't. U. S. Coast Survey.—With Diagrams.

(Communicated by authority of the Treasury Department to the American Associ ation for the Advancement of Science.)

The results of the explorations of the Gulf Stream in the survey of the Coast, have been communicated to the Association. from time to time, as phenomena of peculiar interest have been developed.

The original plan of these explorations having been carefully studied, and having proved successful, has steadily been adhered to. The more recent observations have been directed to that part of the stream between Havana and Cape Florida, known as the channel and strait of Florida.

I have now to present four sections showing the depth and temperature in this most important region of the Gulf Stream. These results are from the observations of Commander B. F. Sands and Lieut. Commanding T. A. Craven, U. S. Navy, Assistants in the Coast Survey, whose names have already been mentioned before the Association in connection with explora

The old spherical musket ball of 69 calibre weighed 340 grains. The new "rifle musket" ball of 58 calibre weighs 500 grains, while still heavier balls are in use in Europe. Thus we see in use balls surpassing in weight the musket ball, and having twice, and even three times the weight of the Swiss ball described in the text, admitted to be, practically, the best in Europe.

tions of the Gulf Stream, and furnish a sufficient guaranty that the results have all the reliability which care, experience and zealous labor could give them.

Section No. 1, from Cape Florida to Bemini was run by Lieut. Commanding Craven, in May, 1855; Section No. 4, by Commander Sands, in May, 1858; and Sections 2 and 3 by Lieut. Commanding Craven, in April and May of the present year, (1859). Sections 2, 3 and 4 are perpendicular to the direction of the Stream at distances of about fifty, one hundred, and two hundred miles from Cape Florida. The Florida strait is funnelshaped, being about ninety miles wide at Havana and about fortyfive miles wide at Cape Florida, the narrowest part.

Form of bottom.-The area of the water way and the form of the bottom are represented on diagrams 7, 8, 9 and 10. The Arabic numerals at the top represent distances from the Florida coast (the Keys) in miles, and the Roman numerals, the positions at which observations are made. The numbers at the left hand represent the depth in fathoms.

Commencing at the Cape Florida section, it will be seen that there is a rapid descent of the bottom to the Havana section, from three hundred and fifty fathoms to eight hundred fathoms, or twenty-seven hundred feet in a distance of two hundred miles. The most shallow as well as the narrowest part of the Stream is therefore at Cape Florida. The deepest water follows the coast of Cuba and the Grand Banks, the depth being eight hundred fathoms at a distance of only five miles from Havana, nearly four hundred fathoms within five miles of Salt Key Bank, and three hundred fathoms close to the island of Bemini. The descent from the Florida side is for the most part gradual, but from the opposite side abrupt. This effect seems to have been produced by the action of the sub-current in wearing a deeper channel upon the concave side of the Stream. At IIavana there is an abrupt descent of nearly a mile within five miles of the shore, while on the side of the Tortugas and Key West the water is comparatively shallow, and the descent gradual. This fact goes to confirm the conclusion that the stronger current of the Gulf Stream makes the circuit of the Gulf of Mexico, since, if it impinged directly upon the island of Key west and the Tortugas we should find its effects in the wearing of a deeper channel on that side.

TEMPERATURES.

Change of temperature with depth.-In a former communication the law of change of temperature with depth was discussed, and types of the curves representing the law were given for different parts of the Stream. The curves were all merely modifications of a more general form. Thus, the cold water between the Gulf Stream and the Coast gave one form; the axis of the Stream

another; and the water beyond the axis a third form, while in the strait of Florida a fourth was developed. It would be natural to expect in the course of many years' explorations by dif ferent individuals with different instruments not even of the same class, that general phenomena of this character should present some contradictions and some inexplicable results. Experience however has confirmed the first conclusions and the constancy of the phenomena. It is not difficult, having the curve representing the temperatures at any position from the surface to the depth of several hundred fathoms to determine from the temperatures alone, in what part of the Stream they were taken.

Temperature in a direction perpendicular to the Stream.-Diagrams 2, 3, 4 and 5, show the changes of temperature for the same depth in each of the sections, and diagrams 7, 8, 9, and 10, the depth for the same temperature.

Bands of warm and cool water.-In the section from Cape Florida to Bemini, the division of the Stream into bands is plainly exhibited, though more faintly than in the northern sections, and the form of the bottom in this section shows the elevations and depressions corresponding to the divisions. In the sections south of Cape Florida, all traces of the bands seem to disappear as well as the ridges of the bottom. The bands therefore seem to have their origin near Cape Florida, and the conclusion stated some years ago, as the probable one, is strengthened, that they are caused by the ridges and valleys of the bottom parallel to the general course of the Stream, and along which the Stream and counter stream have their course.

The Cold Wall.-The cold wall, as an exception to the remark made above in reference to the bands, is traced as far as the Tortugas, and is plainly shown in all the sections with more or less distinctness. In the Sombrero Key section (No. 3) it is strongly marked at depths ranging from seventy to a hundred fathoms, while in all the sections the warm water at the surface overflows the cold wall and reaches quite to the shore.

Diagram No. 6 represents the comparative curves of the cold wall in different sections of the Gulf Stream, including those in the Straits of Florida. The figures at the top show the distances of the cold wall from the shore in the different sections, and the numbers on the left the degrees of temperature. The curves are drawn for different depths in the several sections, as shown in the notes at the bottom of the diagram. The curves g, h, i, k, rep

resent the cold wall in the four sections under consideration.

Longitudinal Sections.-It has been found very difficult to deduce any satisfactory law for the decrease of surface temperature along the axis of the Stream owing to the variability of the temperature of the water of the regions from whence the Gulf Stream is supplied. Two modes of investigating the subject

SECOND SERIES, VOL. XXIX, No. 80.-MARCH, 1860.