April 27, 1876] 



NATURE 



517 



species into the Lower Valais, while others of later origin 

 were principally introduced by human agency. 



During these changes the Mont Blanc district and the 

 country between the Alps and the Jura were still ice- 

 bound, and seeds carried by the wind from the south and 

 west would fall on snow or sterile moraines. And when 

 in their turn these districts were released, their oppor- 

 tunity of being stocked by the flora fast disappearing 

 from the lower levels had gone. The asylums which 

 were earliest opened were most richly supplied and have 

 remained so. 



M. De Candolle considers that a potent cause of the 

 extermination of this flora has been the destruction of 

 the forests which has rendered the climate south of the 

 Alps hotter and drier in summer, and colder in winter. 



The rare plants of the Italian Alps are the remains 

 therefore of an ancient flora like that of St. Helena on its 

 last legs. The climate of Europe tends to become drier, 

 and M. De Candolle thinks it probable that in the course 

 of centuries the centre of Switzerland may in turn become 

 relatively rich in rare species, while the southern slopes 

 of the Alps become poor. In the Lebanon and the Pyre- 

 nees this reversal of conditions has actually taken place, 

 and their southern face — once rich probably in species re- 

 migrating northwards— is now actually poorer than the 

 northern. The Caucasus and the Himalaya are, however, 

 at present comparable with the Alps. T. D. 



DEEP-SEA TELEGRAPH CABLES: HOW 

 THEY ARE TESTED 



THE "testing" of a telegraph cable, whether long or 

 short, proceeds upon the principle that the materials 

 offer to the electrical current a certain resistance : the 

 testing of a cable is the measurement of this resistance. 

 In any cable there are two kinds of these resistance 

 measurements ; one of the resistance which opposes the 

 current in its progress along the conducting wire, the other 

 of that which opposes its lateral dispersion. The conduc- 

 tor-resistance is technically termed the copper-resistance, 

 and is extremely small compared with the other resist- 

 ance. The lateral resistance to the escape of the current 

 opposed by the insulating substance which surrounds the 

 copper-conductor is technically termed the insulation- 

 resistance. Where the resistance to the direct propa- 

 gation of the electric current through a conducting wire is 

 represented in units, the resistance to lateral dispersion 

 through the insulator will be represented by hundreds, or 

 even thousands of millions, of these units. A third pro- 

 perty is that known as the electro-static, or inductive 

 capacity, or simply " charge " ^ of the cable ; in other 

 words, that measured quantity of electricity which the 

 given cable will take up in a given time. So much for 

 the necessary explanation of technical terms. 



The copper-resistance (i), the insulation- resistance (2), 

 and the " capacity " (3) are the three points to be ascer- 

 tained in the testing of a cable ; and it is useful to inquire 

 why these are the points to be ascertained. 



The chief commercial requisite in any cable, and upon 

 which depends its value to its owners, is the speed with 

 which signals can be transmitted. Speed^depends directly 

 upon two of the foregoing points (that is upon the copper 

 resistance and " capacity"), and indirectly upon the insula- 

 tion-resistance. Popular assumption is very much given to 

 the idea that the electrical worth of a cable increases with 

 its insulation-resistance ; as usual with popular notions 

 this is only half-truth. That the cost of a cable follows 

 the ratio may or may not be, but it is certain that above a 

 definite limit the thickness of the insulating coating has 

 no effect upon the practical working condition of the 



"Capacity" and "charge" are not equivalent terms, although they are 

 so considered m this article to prevent confusion, by the general reader, 

 with the ordinary meaning of the word " capacity." The capacity of a 



I ''*?'^'°-'^ constant, .while the charge varies with the battery power 



cable. It may be that minor indirect benefits arise, but 

 with these, under the present consideration of the prac- 

 tical testing of a cable, we have nothing to do. A certain 

 standard of insulation-resistance attained, there remain 

 the two points, first, of the resistance offered by the cop- 

 per wire ; secondly, of the charge. Now it is collaterally 

 to be understood that, as there can be obtained through a 

 pipe a greater flow of liquid when the pipe offers little 

 resistance to the flow, so through the conductor of a cable 

 can a greater flow be obtained when the conductor has 

 low resistance. With most of the Atlantic cables each 

 nautical mile of the conductor has a resistance equal to 

 that of three to four of the arbitrary units selected by the 

 profession for comparison. There are in use two units of 

 electrical resistance, namely, that determined by a com- 

 mittee of the British Association and the Siemens unit. 

 These units are very nearly of the same value, one 

 Siemens' mercury unit (the resistance offered by a column 

 of pure mercury of one metre length and one square milli- 

 metre section at 0° C.) being equal to o'9536 of an Ohm, the 

 technical term for a British Association unit. There is, 

 then, to be considered an electrical length as well as an 

 absolute (or ordinary) length ; the proportion that one 

 bears to the other being known, the measures are con- 

 vertible. Vague as may appear to the reader this idea of 

 electrical resistance, when he knows that of a copper wire 

 of given diameter or weight two lengths offer twice the 

 resistance of one, he is as learned as the most skilled elec- 

 trician who virtually knows no more. 



The consideration of the electrical capacity of a cable 

 is more difficult. While the two other points relate to 

 mass, the question of capacity involves that of surface, 

 and of a property of the insulating material of the 

 cable known as its " specific inductive capacity." The 

 material with which long telegraph cables are insulated is 

 gutta-percha. Two different cables may be insulated with 

 this material to precisely the same dimensions, both as re- 

 gards the thickness of the insulator and the thickness of the 

 copper wire, but the " charge " taken by these cables may 

 be very different, and the difference wiU be due to difference 

 in the specific facilities offered by the two gutta-perchas to 

 induction. This difference between various kinds of gutta- 

 percha is as inherent as is the difference between resistances 

 to conduction offered by different metallic alloys, and is 

 probably very often due to want of homogeneity of the 

 substance. It is by judicious selection and careful mani- 

 pulation that the cable manufacturer is enabled to main- 

 tain a certain standard for any particular cable in question. 

 Capacity, however, not only varies with the insulating ma- 

 terial, but it also varies with the amount of surface of the 

 conductor. It is different with different thicknesses of 

 insulating material, but in this respect, after a certain limit 

 has been passed, the decrease in capacity is very small for 

 very large increase in the thickness of the insulating 

 material. 



High charge is incompatible with high speed. That cable 

 will, other conditions the same, have the greatest speed 

 in which the charge, or the fraction of the charge to be 

 altered at each signal, is least. Professional necessity has 

 given rise to a unit of quantity of electricity termed a 

 "farad," of which the "microfarad" is the millionth 

 part. The capacity of a telegraph-cable generally ranges 

 from three to four-tenths of a micro-farad per nautical 

 mile. 



The object of testing a cable is, then, to ascertain whether 

 the insulation reaches the amount specified, and whether 

 the conductor-resistance and the charge are of the re- 

 quired minimum. As these tests are each applied sepa- 

 rately to the cable, their consideration will fall under the 

 several heads. It would clearly be impossible within the 

 limits of this paper to describe the many methods which 

 have from time to time been proposed and in use for the 

 testing of telegraph cables. The first methods of testing 

 submarine lines are undoubtedly due to Dr. Werner 



