44 



THE INDIA RUBBER WORLD 



[November i, 1904. 



oxide of zinc, has a dielectric strength comparable to that of 

 pure vulcanized Pard { 1 5 to 20 kilovolts per millimeter) ; and a 

 specific inductive capacity of 4 to 4.2. A rubber with 64 per 

 cent. Para, S per cent, sulphur, 16 per cent, talc, 8 per cent, 

 minium, 4 per cent, oxide of zinc has about the same dielec- 

 tric strength as above mentioned, while its specific capacity 

 reaches five. A rubber largely loaded with sulphur and talc' 

 for example, Para 100, talc 40, and sulphur 40, has a capacity 

 as high as 6.10, with a dielectric strength of the same order of 

 magnitude as before. A mixture of Pard 40, carbonate of lime 

 45, sulphur 5. has a standard inductive capacity of 4.6. Very 

 large variations of capacity, accompanied by high dielectric 

 strength, are obtained by loading rubber with more or less sul- 

 phur and golden sulphurate of antimony still remaining first- 

 class rubber. Much larger capacitier., 10 to 12, are to be ob- 

 tained, of course, by using very large percentages of India-rub- 

 ber substitutes, such as gypsum, lime, baryta, etc. ; but we then 

 arrive at inferior classes of rubber, which have not a dielectric 

 strength to be compared with the above-mentioned combina- 

 tions. 



» • » 



It is very easy to manufacture rubber cables with layers dis- 

 posed in the order of decreased specific capacity, from the 

 center to the circumference. These cables will aflord a more 

 uniform gradient to an alternating current, and hence more 

 safety, with equal thickness. By using paper on the rubber, as 

 above explained, we concentrate the more costly rubber insula- 

 tion in the inmost part of the cable, where its higher specific 

 strength is actually utilized. 



A sahiple of a single core cable made by this method for 50- 

 kilovolt effective tension, between the copper and the outer 

 sheathing, has the following specifications: Conductor, 19-wire 

 strand, each wire 3,3 millimeters diameter ; section of copper 

 162 square millimeters. The strand is put in a lead tube hav- 

 ing 18 millimeters outer diameter. It is insulated with a 

 first layer of rubber, 2.5 millimeters thick, having a specific in- 

 ductive capacity of 6.1 ; then with a second and a third layer of 

 rubber of respectively 2.3 and 4.5 millimeters thick and 47 to 

 4.2 standard inductive capacity. On the rubber there is a 

 layer of impregnated paper 5.2 millimeters thick, having a 

 standard jnductive capacity of 4. The cable is then lead-cov- 

 ered. The total thickness of insulation is 14.5 millimeters. 



At 50,000 volts, the maximum strain in the first layer of rub- 

 ber is 4400 volts per millimeter; in the second layer it is 445° 

 volts, in the third 4150 and in the paper 3250 volts per milli- 

 meter. With a homogeneous dielectric, the maximum strain 

 would be 5800 volts. This cable was tested for one hour at 

 each of the following voltages : 35,000 effective volts, 40,000, 

 45,000, 50,000, 55,000, 60,000, 65,000, 70.000, 75,000, So,ooo, 85,- 

 000, 90,000, 95,000, and four hours at 100,000 volts without per- 

 foration. After the 80,000 volts test, its temperature was a few 

 degrees higher than that of the room ; and after four hours at 

 100,000 volts, twenty degrees centigrade higher. 



* * * 



Gutta-percha possesses also very great dielectric strength, 

 comparable to that of good rubber, 15 to 20 kilovolts per milli- 

 meter. It is not used for insulating cables for lighting or 

 power purposes, because of its very high price, and also its 

 especially low melting point. Such cables can easily reach a 

 temperature which softens Gutta-percha. A possible applica- 

 tion of Gutta-percha is for cables crossing lakes, rivers, and, 

 generally speaking, for laying in cold water. It is then advis- 

 able to make a first layer of rubber insulation, on which Gutta- 

 percha is laid so that the latter, being in contact with external 

 cold water, can not heat very much. Many manufacturers do 



not trust the impermeability of rubber cables, and this external 

 coat of Gutta-percha, absolutely waterproof, adds its own di- 

 electric strength to that of rubber and obviates the inconven- 

 ience of having a heavy lead pipe, as employed by the manu- 

 facturers to which I have alluded. It is often advisable in such 

 cables to avoid splices, and for the sake of facility of transport 

 and laying, they can be single cored, rather than three cored. 

 I may add that single core cables for very high tensions, re- 

 quiring generally a low current strength, can often be armored 

 with steel wires; the steel wires can be separately wrapped 

 with tarred manila, in order to lessen the section of the metal 

 and increase the magnetic and electric resistance of the cross 

 circuit. For example, a 2.5 millimeter steel wire wrapped to 5 

 or 6 millimeters with manila, may be used without any great 

 inconvenience from hysteresis or self-induction ; the drop of 

 pressure by self-induction can have in such cables no more im- 

 portance than the drop by ohmic resistance. 



» » ♦ 



I WOULD like to add something on the properties of various in- 

 sulating materials. These materials are influenced by Rontgen 

 rays, which lessen their specific insulation and perhaps also their 

 dielectric strength. But cables are not made to be submitted to 

 such rays, although they often experience brush discharges and 

 some other emanations, which may have similar influences. I 

 should like to add that temperature lessens the resistance of 

 the insulation very quickly, as expressed in megohms. A pa- 

 per cable at 35' Centigrade shows but one-thirtieth of the me- 

 gohms it has at 15° Centigrade. But temperature has very little 

 influeuce upon strength to resist breakdown. Palm oil melted 

 at 50° Centigrade gives a strength corresponding to that of 

 the best oils for transformers at ordinary temperature. I have 

 drawn experimental curves of dielectric strength of melted par- 

 affin at 55 Centigrade and at 85° Centigrade from 10 up to 160 

 kilovolts; they are very similar. This allows us to conclude 

 that in this respect cables cannot differ very much. I have 

 tested two reels of paper cables, each cut in five pieces, im- 

 mersed in baths at zero, 1 5', 35°, 70°, and 100° Centigrade. The 

 dielectric strength did not lessen by raising temperature, per- 

 haps at zero it was less than at 70°. I noted in some oils some- 

 thing similar, but dielectric strength is too complex a phenom- 

 enon to be discussed on small experimental differences. Of 

 course, that cannot justify us in working at high tensions with 

 cables too highly heated, for it is probable that heat would fa- 

 cilitate a chemical decay of the dielectric; but a momentary 

 elevation of temperature is not so much to be feared as one 

 would think at first sight. 



THE WEARING OF RUBBER COLLARS. 



THE New York Press says that Mr. Duke, " worth millions, 

 all of which he made out of tobacco, is not the only 

 person in the world who wears rubber collars. They are be- 

 coming popular with schoolboys on account of their economy. 

 Seven linen collars a week cost to launder fourteen cents. A 

 rubber collar, price thirty-five cents, can be cleaned without 

 trouble every morning, or a dozen times a day, and will last 

 three months. As an experiment, I tried one of these collars 

 on a fishing trip and it was not a success. Being impervious, it 

 caused the neck to sweat too freely and kept the neckband of 

 the shirt wringing wet. Others have had the same experience. 

 For ten years the manufacturers have moved heaven and earth 

 in vain to induce the trade to handle these rubber collars. 

 Their business is confined to two small shops in this city. Per- 

 haps if they would judiciously advertise, something might be 

 accomplished. They might build up a mail order business." 



