they were shells at all! These minute creatures live near the surface and their shells have been rained down, so to speak, for ages. When specimens of the soundings brought up are examined, they are found to be similar to the material forming our chalk cliffs, which have no doubt been similarly built ages ago. As these microscopic shells were so fragile that a breath would almost destroy them, they afforded a proof that there were no currents at the bottom moving over the surface of this plateau; for had the shells been rolled to and fro, their delicate organism would have been destroyed.

An idea very generally prevailed that the ocean was practically unfathomable; that is to say, that, owing to the pressure of water, nothing could possibly sink to the bottom, and that sooner or later everything-even a cannon-ballwould find itself in a state of equilibrium and descend no further. Science had, however, shown that owing to the water itself being less compressible than even metals, its specific gravity would not be increased at great depths by the pressure of the column of water above, to so great an extent as the weight used for sounding, of whatever material it was composed. It therefore followed that at any given depth the sounding lead would be relatively heavier as compared with the water around it, than when at the surface; and hence the lower it went the greater would be its tendency to sink.

The sounding apparatus used was of a very ingenious kind, and its arrangement will be best understood by a reference to fig. 1.

A rod of iron (B) is attached to the sounding-line, at the bottom of which are fixed a few quills with their ends open. This rod is passed through a hole in the centre of a cannonball (4), which can move loosely upon the rod, but is held in its position by a cord passed round it and fastened at each end by loops to two curved arms attached to the soundingline. On the line reaching the bottom, the weight of the

cannon-ball drives the end of the rod into the sand. The curved arms fall down and release the loops holding the cannon-ball. On the line being hauled up, the rod then passes clear, leaving the cannon-ball at the bottom, and carrying to the surface the quills containing some of the sand or ooze from below.

In constructing the first cable, in 1857, it was considered necessary that it should be capable of sustaining five or six miles of its own weight in water, when suspended vertically; so as to allow of laying to, if required, during submersion. At the same time the cable had to be heavy enough to draw itself freely from the hold, somewhat in excess of the ship's speed, and to sink readily; so as to avoid lashing of the waves in rough weather, and to pass without interference through the currents near the surface. After experiments on upwards of sixty kinds of cables, one was selected with a central conducting strand composed of seven small copper wires and coated with three distinct layers of gutta-percha. This core was then surrounded with tarred yarn, and covered over with eighteen strands of iron wire, for an outer protection, as shown in fig. 2.

The object of using a bundle of wires instead of a single solid conductor, was to prevent a flaw in one of the wires, at any point, interfering with the conductivity of the cable; as the electricity could in such case pass along the remaining wires of the strand without interference. In the same way three coatings of gutta percha were applied successively, so that any minute air hole or other defect in one layer would be covered up by the other layers. The chance was excessively small of a defect in manufacture occurring in each of the three layers at precisely the same point! It is well known that this cable when laid in 1858 was worked for a month, and then communication ceased; owing to the gutta percha insulation becoming defective at a fault, which the tests

shewed to be about two hundred and seventy miles from Valentia. The electrical leakage through the fault had been augmented by the strong currents used in passing signals.

This failure discouraged further prosecution of the enterprise for some years; but the experience gained by it was of the greatest importance, and really formed the germ of the more permanent success achieved this year. It was seen that with deep sea cables it was advisable to construct them proportionately stronger and specifically lighter than the first Atlantic line, so that they might be recoverable at great depths. It was also obvious that for so long an unbroken circuit the conductor should be larger and the gutta percha insulation more perfect, so as to enable a greater speed of transmission to be attained with a less intense current; in fact, the weaker the electric charge capable of producing an effect at the other end, the less tendency it would have to burst its way through the gutta percha at any defective point, and, therefore, the more likely the cable would be to last.

In 1865 a new Atlantic cable was made, which was a great improvement in many respects upon its predecessor. The outer protecting strands were formed of a combination of iron wires cased with hemp, saturated with a tarry compound as a protection from rust. The copper conducting wire consisted of seven strands as before; but weighed three hundred pounds per mile, or nearly three times that of the 1858 conductor. It was insulated by no less than eight coatings of gutta percha and a viscous insulating compound laid alternately over one another. The wire had thus not only three times the conducting power, but a far better insulation than its predecessor; and was capable of passing seventeen words per minute, while the former cable could only transmit three or four words per minute. The cable as a whole was specifically lighter in water and far stronger. It weighed only fourteen hundredweight in water, while the old

cable weighed thirteen hundredweight; but it could withstand the strain of seven tons and a half, while the strength of the 1858 cable was only about half as great. The difference will be better understood by a reference to fig. 3.

As the risk to submarine cables when laid is in a great measure confined to the shallow water near each shore, where there is the chance of damage from anchors, fishing trawls &c., the shore ends of the Atlantic Cable, for a distance of thirty miles out, were made of an exceedingly massive form. Their weight is no less than twenty tons per mile, and they are protected by an outer spiral casing of iron rods; so that, even if accidentally caught by the anchor of a line-of-battle ship, they would probably hold it without suffering injury.

The weight and bulk of the cable being so enormous, when multiplied by the length to be made (twenty-three hundred miles) it was determined to engage the Great Eastern, which was then seeking employment almost in vain. By this arrangement the whole of the cable could be stowed in one ship, while without her aid four ships of the largest size would have scarcely sufficed; and, as in 1858, the cable would have had to be much smaller in size. Even with present experience it would be a most dangerous experiment to attempt to lay a cable piecemeal across the Atlantic from a series of vessels; for rough weather might at any time prevent the ends being successively joined, as each ship finished its portion of the task.

The mission of this vast ship was at last discovered, and she was speedily prepared for the work. Huge tanks were built within her to receive the cable, and keep it continually saturated with water; so that in case the slightest fault occurred prior to the insulated cord passing into the sea, it would be at once detected.

I now come to the machinery devised by Messrs. Canning and Clifford for laying the cables of 1865-6,