252 
THE TEACHING OE MECHANICS BY EXPERIMENT. 
There is of course nothing new in this notion. It is as old as the 
study of mechanics itself. Galilee and Newton did not disdain the 
aid of experiment in fixing their mechanical concepts : each of them 
describes experiments which had this intention. And to go further 
back we all know how much Archimedes owed, and admitted he owed, 
to an experiment while he was struggling to solve the problem of 
finding the specific gravity of an irregular solid. He was able to cry 
tvprjKa only when he made the simple experiment of taking a bath. 
The advantage of experiment has been present to the mind of many 
a teacher. That tutor of Trinity who flourished a century ago 
and whose name is immortalised as the inventor of “ Atwood/s 
Machine ” must have been conscious that the equation which ex¬ 
presses the motion of a body under uniform acceleration gains 
something of meaning and point by such a concrete presentation. 
And to take a more recent example, Sir Kobert Ball has shewn in his 
book of Lectures on Experimental Mechanics how much may be done 
by means of quantitative experiments performed on the lecture table 
to make the subject intelligible to people with little or no mathemati¬ 
cal training. 
Of late years the idea has been steadily gaining ground that 
even a systematic and more or less mathematical exposition of 
mechanics gains much when it is accompanied or followed by experi¬ 
ments which the students perform themselves. This has led to the 
creation of mechanical laboratories and to the addition of experiments 
in mechanics to the work of existing laboratories of physics or of 
engineering. In some cases, as at the Central Technical College, 
where Professor Henrici has done much to show the value of this kind 
of teaching, the mechanical laboratory is separate. In other cases it has 
grown up within and forms part of a laboratory of engineering. A 
notable example is to be found in the City and Guilds Techuical Col¬ 
lege at Finsbury. The work done there by Professor Perry has been 
of a peculiarly important and suggestive kind and it is to him perhaps 
more than to any other teacher that we owe our present knowledge of 
what may be done and what should be done in the teaching of 
mechanics by experiment. I have myself found his example con¬ 
tagious, and have to acknowledge receiving much help from the little 
book on “Practical Mechanics” in which he originally expounded 
his method of teaching. No writer insists more strongly on the value 
of a constant appeal to experience and on the advantage of quantita¬ 
tive experiment as a means of acquiring experience. 
In the new engineering laboratory at Cambridge mechanical experi¬ 
ments take a prominent place, and we have not hesitated to copy or 
adapt from any quarter whatever seemed likely to form a useful addition 
to the stock of experimental apparatus. My purpose in this lecture is to 
give you some idea of the scope of such experiments and to bring 
before you a few representative examples of the apparatus with which 
they are carried out. The first thing a student has to do in a mechan¬ 
ical laboratory is to learn something of the art of measurement, the 
measurement especially of mass and time, of length and area. The 
