_The kinks are caused by seismic disturbances, and they show how much distortion a cable can suffer and still remain in good electrical condition, as this was found to be._] And this brings us to the consideration of one of the greatest practical advantages of the seismograph--in the exact location of cable breaks. Indeed, a large proportion of these breaks are the result of earthquakes. In a recent report Professor Milne says that there are now about twenty-seven breaks a year for 10,000 miles of cable in active use. Most of these are very costly, fifteen breaks in the Atlantic cable between 1884 and 1894 having cost the companies $3,000,000, to say nothing of loss of time. And twice it has happened in Australia (in 1880 and 1888) that the whole island has been thrown into excitement and alarm, the reserves being called out, and other measures taken, because the sudden breaking of cable connections with the outside world has led to the belief that military operations against the country were preparing by some foreign power. A Milne pendulum at Sydney or Adelaide would have made it plain in a moment that the whole trouble was due to a submarine earthquake occurring at such a time and such a place. As it was, Australia had to wait in a fever of suspense (in one case there was a delay of nineteen days) until steamers arriving brought assurances that neither Russia nor any other possibly unfriendly power had begun hostilities by tearing up the cables. There have been submarine earthquakes in the Tuscarora, like that of June 15, 1896, that have shaken the earth from pole to pole; and more than once different cables from Java have been broken simultaneously, as in 1890, when the three cables to Australia snapped in a moment. And the great majority of breaks in the North Atlantic cables have occurred in the Newfoundland hollow, where there are two slopes, one dropping from 708 to 2,400 fathoms in a distance of sixty miles, and the other from 275 to 1,946 fathoms within thirty miles. On October 4, 1884, three cables, lying about ten miles apart, broke simultaneously at the spot. The significance of such breaks is greater when the fact is borne in mind that cables frequently lie uninjured for many years on the great level plains of the ocean bed, where seismic disturbances are infrequent. The two chief causes of submarine earthquakes are landslides, where enormous masses of earth plunge from a higher to a lower level, and in so doing crush down upon the cable, and "faults," that is, subsidences of great areas, which occur on land as well as at the bottom of the sea, and which in the latter case may drag down imbedded cables with them. It is in establishing the place and times of these breaks that Professor Milne's instruments have their greatest practical value; scientifically no one can yet calculate their value. [Illustration: Record Made on a Stationary Surface by the Vibrations of the Japanese Earthquake of July 19, 1891.