RE'S ID UAL PHENOMENA. 107 



shall be chosen? That which is clear and definite, and from which 

 results can be deduced in a form which permits of their being tested by 

 experiment. 



If such an hypothesis be found, it then becomes necessary to ask, 

 Does this hypothesis explain facts other than those included in the 

 special residual phenomenon under consideration ? An hypothesis which 

 explains, or seems to explain, an isolated phenomenon, but which does 

 not include other phenomena within its grasp, or which does not leap to 

 the discovery of hitherto unknown facts, may be a true hypothesis, but 

 it is certainly one which must be accepted with caution, and only pro- 

 visionally until a better be found. 



Finally, the new hypothesis must be in keeping with the well-estab- 

 lished laws of nature. An hypothesis which contradicts any of these 

 can not be accepted, although it may explain the special phenomenon to 

 give a reason for which it has been called into existence. 



The recent history of natural science furnishes many examples of 

 the use of residual phenomena. Let me mention two only : one, in 

 which an hypothesis has been suggested, proved, and adopted ; another, 

 in which the value of the hypothesis suggested is not yet finally deter- 

 mined. 



It is well known that plants derive their support from the air and 

 the soil ; that support consists partly of mineral, partly of vegetable 

 matter. But the curious fact was noticed that the leaves of certain 

 plants frequently had adhering to them remains of insects or even entire 

 insects. Following up this fact, Mr. Darwin and others have estab- 

 lished the generalization that members of more than one species of 

 plants derive their nourishment mainly from animal matter, and that 

 these plants thrive better upon such food than upon the ordinary kinds 

 of plant-food. Thus another thread has been added to the bond which 

 visibly connects the animal and vegetable kingdoms. 



The chemical elements have long been regarded as truly elementary 

 bodies, that is, as bodies from which no form of matter other than 

 themselves can be obtained. But phenomena presented by the spectra 

 of certain of these elements seem almost inexplicable by the commonly 

 accepted view. Mr. Lockyer has carefully examined many of the so- 

 called elementary spectra, at temperatures varying from that of a gas- 

 flame to that of the star Sirius, and, in order to explain the phenomena 

 noticed, he has provisionally adopted the hypothesis that the so-called 

 elements are really compound bodies. This hypothesis, whether even- 

 tually confirmed or refuted, suggests a large field for research to the 

 chemist and to the physicist, from which neither can fail to reap most 

 valuable results. 



The observed residual phenomena of nature which yet await solution 

 are many and varied ; every branch of scientific work presents its own 

 list. Let me glance at a few, and they shall be chiefly chosen from 

 those phenomena which are investigated by the science of chemistry. 



