Mat 5, 1911] 



SCIENCE 



675 



merely operating, their work rises witli 

 some higher power of the number than the 

 first power. It approaches the square for 

 two men and the cube for three. Two men 

 cooperating with two different and special 

 pieces of apparatus, say a special furnace 

 and a pyrometer, or an hydraulic press and 

 new chemical substances, are more power- 

 ful than their arithmetical sum. These 

 facts doubtless assist as assets of a research 

 laboratory. 



When a central organization, such as a 

 laboratory, has access to all parts of a large 

 manufacturing plant and is forced sooner 

 or later to come into contact with the vari- 

 ous processes and problems, the various 

 possibilities and appliances, it can hardly 

 fail to apply, in some degree, the above law 

 of powers. 



As a possible means of illustrating the 

 almost certain assistance which one part 

 of a manufacturing plant may give another 

 when they are connected by experimenting 

 departments or research laboratories, and 

 how one thread of work starts another, I 

 will briefly review part of a single fairly 

 connected line of work in our laboratory. 

 In 1901 the meter department wanted 

 electrically conducting rods of a million 

 ohms resistance. These were to be one 

 quarter inch diameter by one inch length. 

 In connection with this work we had to be- 

 come fairly familiar with published at- 

 tempts at making any type of such high 

 resistances. Some kind of porcelain body 

 containing a very little conducting mater- 

 ial seemed a fair starting formula after the 

 resistance of almost all kinds of materials 

 had been considered. Our own porcelain 

 department was of a great help in showing 

 us how to get a good start. We learned 

 how and what to mix to get a fair porcel- 

 ain, and we found that small quantities of 

 carboxTindum or of graphite would give us 

 the desired resistance about once in a hun- 



dred trials. The rods could be made, but 

 the variation of their resistance when 

 taken from the porcelain kUn and when 

 they were made as nearly alike as we could 

 make them, was often so many thousand 

 fold that something new had to be done to 

 make a practical success. A small electric 

 furnace was then devised for baking the 

 rods and this was so arranged that the rate 

 of rise of temperature, the maximum 

 temperature reached and the duration of 

 heat at any temperature was under control 

 and was also recorded. The desired result 

 was obtained and this work was thus fin- 

 ished. It gave us a certain stock of 

 knowledge and assurance. 



At that time a very similar problem was 

 bothering one of the engineering depart- 

 ments. Lightning arrester rods, part of 

 the apparatus for protecting power lines 

 from lightning, were needed. Their di- 

 mensions were | X 6 inches and they 

 needed to have a definite but, in this case, 

 low resistance, and could apparently not be 

 baked in a porcelain kiln. The necessary 

 variations in such a kiln are so great that 

 in practise many thousand rods were re- 

 peatedly fired and afterward tested to 

 yield a few hundred of satisfactory prod- 

 uct. All the cost of making an entire batch 

 would have to be charged against the few 

 units which might be foiind satisfactory, 

 and in many eases there were none good in 

 a thousand tested. It was evident that 

 regulation and control of temperature was 

 necessary. This was found to be imprac- 

 ticable in case any considerable niunber 

 were to be fired at one time, as the heated 

 mass was so great that the rods near the 

 walls of the retort received a very differ- 

 ent heat treatment from those near the 

 middle and were consequently electrically 

 different. This was still the case even 

 when electrically heated mufSes were used. 

 This difficulty led to experiments along the 



