no THE POPULAR SCIENCE MONTHLY 



Pickering has now been for about thirty-seven years the head of the 

 astronomical observatory at Harvard. The present physical laboratory 

 at Harvard was built in 1884. Of late years laboratories have been built 

 at all our colleges, and there has developed a tendency to make them very 

 large and costly. Two of the latest, the Palmer Laboratory at Princeton 

 and the Sloane Laboratory at Yale, have gone well beyond the quarter 

 of a million mark. The largest and best equipped laboratory in the 

 country is that belonging to the national government, and known as the 

 Bureau of Standards, which, in its brief history of about ten years, has 

 had over two millions of dollars spent upon it. When we consider that 

 this institution is entirely separated from teaching, we must believe that 

 work of great importance is done there to justify this great outlay. 

 Permit me to describe what some of the functions of such a laboratory 

 are, and incidentally to explain some of these devices that are to be 

 found in any great modern laboratory. 



At the Bureau of Standards we find five large buildings, each de- 

 voted to a particular purpose. These have cost $712,000. In the largest 

 we find the divisions of weights and measures, of heat, and of light. 

 The chief objects of the Bureau of Standards being necessarily practical, 

 the researches undertaken there are limited in scope by this considera- 

 tion. Nothing is perhaps more practical than the verification of the 

 standards of weight by which commodities are bought and sold. Even 

 the weights of the mint are tested at the Bureau of Standards. Accord- 

 ingly in a basement room mounted upon heavy brick piers, which are a 

 prominent feature in every physical laboratory in order to secure free- 

 dom from vibration, we find extremely accurate balances, some of which 

 are capable of weighing a body with an accuracy of one part in fifteen 

 or twenty millions. The comparison of two equal weights is probably 

 susceptible of greater accuracy than any other physical operation. It is 

 to be remarked that in order to attain this degree of accuracy the bal- 

 ance has to be operated in vacuo, the whole instrument being placed 

 in a case from which the air is pumped out, and all operations of trans- 

 ferring the weights being conducted from a distance by means of con- 

 trolling rods or shafts, since the heat of the observer's body near the 

 balance would so change the length of the beam as to render such an 

 accuracy impossible. 



Next to weighing comes the measurement of length, which is sus- 

 ceptible of about the same accuracy. Here again, the effect of changes 

 of temperature in causing metal scales to expand has to be provided 

 against, so that the work has to be carried on in a subterranean vault, 

 where the changes of temperature are made as small as possible. In the 

 division of heat great practical importance belongs to the measurement 

 of temperatures. Thousands of thermometers of all sorts are sent here 

 yearly to be tested. Before the existence of the Bureau of Standards 



