TBANSACTIONS OF SECTION A. 597 



o>ne ought not to do those things, but still if one wrote much one was very apt at 

 some time or other to make very hasty statements. When the Society of Arts 

 asked him to give a lecture on lightning-conductors he thought to himself, ' Yes, I 

 will tell them about induction,' because he knew that the Lightning-rod Confer- 

 ence had not called attention to that nor yet to magnetic inertia or whatever they 

 called it. He had happened to write a paper which appeared, he thought, in 

 ' Nature ' — he did not know where the quotation made by Mr. Preece came in — and 

 in that paper having referred to the magnetic permeability of the two metals, copper 

 and iron, he had stated that iron was 90,00U times worse than copper. He went 

 on to say, therefore, that although iron was cheaper it was more difficult to melt, 

 Lad a higher specific heat, and in a variety of ways was better than copper, yet in 

 regard to this electro-magnetic inertia it was enormously worse. But before 

 actually giving those lectures, on again making a few experiments, to his surprise, 

 he found that, so far from being worse, it was often rather better — that iron, even a 

 thin wire of iron, carried oif the discharge better than a thick wire of copper. He 

 did not press the point that iron is better than copper, and he had never gaid that 

 copper has more self-induction than iron — that would be a mad thing to say. It 

 might happen for some reason or other that the copper obstructed the current more 

 than the iron, but he preferred to say that they are just equal. 



Now Mr. Preece said that the function of a lightning-conductor is to prevent a 

 flash from striking the conductor ; that is to say, that a lightning-conductor never 

 ought to be struck or it fails. But they are struck because they get melted. Yet 

 at the beginning sentence Mr. Preece said that they never failed. He did not know 

 which of these statements was to be taken as correct. If a lightning-conductor can 

 prevent a flash from occurring by its repellent action well and good, but he had 

 shown in those lectures that there are cases where a point has no protective action 

 whatever, when a point could be struck by a thick and heavy flash. There were 

 other cases where the point acts with a brush or fizz and neutralises the electric 

 charge without a flash. They could not always do it. And so the lightning-rod 

 has two functions, one is to be repellent if it can and the other is to carry off" a flash 

 when it cannot help receiving it. But they must remember that a flash, at least 

 that the electric charge, has a certain amount of energy, and that has to be dis- 

 sipated somehow. It was not a question of a certain quantity of electricity to be 

 conducted to earth and then there was an end of it. There must be a certain 

 amount of energy, they must dissipate it somehow, and thej' could not expect to 

 Jiocus-pocus it out of e.xistence by saying they could conduct it down to the earth. 

 The quicker they tried to conduct it down to the earth the more searching and 

 ramifying disturbances they were likely to get. It might be better to let it trickle 

 down slowly by using a moderately bad conductor than to rush it with extreme 

 vehemence down a good conductor; just as it would be safer to let a heavy 

 weight suspended in a dangerous position down slowly rather than let it drop as 

 quickly as possible. 



Concerning the length of flashes he wished he had any information, but he had 

 none ; it was one of those things which their friends, the meteorologists, must 

 determine for them. It was very important to know the length of a flash. He 

 had found it stated in books that flashes were a mile long and perhaps more. Mr. 

 Preece thought they were only 600 feet long. That was a matter of fact which 

 could be investigated, but of which he had no direct first-hand information. 

 Whether the spark-length is proportional to distance or not he would say that the 

 experiments with oscillating currents to which Mr. Preece referred were conducted 

 with alternating dynamos between points. Now the area of cloud and the area of 

 earth below it are not points, they are flat. 



There the law did not seem to hold, but then it ought not to hold, so that 

 would be all right. But between flat surfaces it ought. The spark-length ought 

 to increase with a ditt'erence of potential between the two flat surfaces. At the 

 eame time if there were points on the earth's surface big enough and which in any 

 fashion could act as points, then of course it would not hold. But the oscillating 

 current did not apply as regarded the length of the spark because until the dis- 

 charge occurs there is no oscillation— it is a mere static charge. It is exactly like 



