6 PROCEEDINGS OF THE ROYAL SOCIETY OF QUEENSLAND. 
of their component particles. Indeed a certain advance 
is made here, passing from simpler to more complex phe- 
nomena, if to these particles they attribute no relationships 
other than those shewn in Nature. But when they take 
on themselves to postuiate arbitrary unknown shapes and 
sizes and doubtful positions and motions, and further to 
invent some oceult fluid, freely permeating bodies, endowed 
with perfect mobility and disturbed by occult motions; 
then are they given over to dreams, to the neglect of the 
real nature of things which, truly, is sought in vain by 
fallacious guesses seeing that it can with difficulty be 
studied by means of the most careful observations. Those 
who base their investigations on hypothesis, even if they 
then proceed most accurately according to the Laws of 
Mechanics, may be said to concoct fables, neat perhaps 
and beautiful but fables nevertheless. 
‘There remains the third school, which concerns itself 
with Experimental Philosophy. Its members attempt to 
derive the causes of all things from the simplest principles 
possible but they admit as a principle nothing that is not 
given by observed phenomena. ‘They do not invent 
hypotheses and introduce them into Physics except as 
propositions into whose truth they must inquire. Their 
method falls into two divisions, one Analytic the other 
Synthetic. The forces of Nature and the simpler laws 
of forces are deduced by analysis of certain selected phe- 
nomena, and from them is found by synthesis the nature 
of the remainder.’’ 
The method outlned in the last paragraph shews the 
main characteristics of modern scientific method; the 
selection of the more outstanding results of observation 
for detached analysis, the formulation of a provisional 
hypothesis as a possible explanation the validity of which 
is to be tested further, the application of results obtained 
from the simpler problems in the building up of a general 
theory covering more complex eases, are all in keeping 
with our present-day views. It shews a marked advance 
on the earlier method of attempting to pass in one bound 
from observation to general theory. 
The need to proceed carefully to the hypothesis after 
preliminary analysis and selection of data had been shown 
by the failure of the old ways; the possibility of so doing 
