INTRODUCTION 



Ideally, the scientific method consists of making careful observations 

 and arranging these observations so as to bring order into the phe- 

 nomena. Then one postulates a hypothesis or conceptual scheme which 

 will explain the facts at hand and make predictions about the results of 

 further experiments or observations. Sciences differ widely in the extent 

 to which prediction is possible, and the biological sciences have been 

 held bv some to be not truly "scientific," for they are not completely 

 predictable. However, even physics, which is generally regarded as the 

 most scientific of the sciences, is far from completely predictable. 



The history of science shows that although many scientists have 

 made their discoveries by folloAving the precepts of the ideal scientific 

 method, there have been occasions on which important and tar-reaching 

 theories have resulted from making incorrect conclusions from erroneous 

 postulates, or from the misinterpretation of an improperly controlled 

 experiment! There are instances in which, in retrospect, it seems clear 

 that all the evidence for the formulation of the correct theory was kno^vn, 

 yet no scientist put the proper two and two together. And there are 

 other instances in which scientists have been able to establish the correct 

 theory despite an abundance of seemingly contradictory evidence. 



In most scientific studies one of the ultimate goals is to explain the 

 cause of some phenomenon, but the hard-and-fast proof that a cause 

 and effect relationship exists between two events is really very difficult 

 to obtain. If the circumstances leading to a certain event always have a 

 certain factor in common in a variety of cases, that factor may be the 

 cause of the event. The difficulty, of course, lies in making sure that the 

 factor under consideration is the only one common to all the cases. It 

 would be wrong, for example, to conclude from the observation that 

 drinking Scotch and soda, bourbon and soda, and rye and soda all 

 produce intoxication, that soda is the only factor in common and there- 

 fore is the cause of the intoxication. This method of discovering the 

 common factor in a series of cases that may be the cause of the event 

 (known as the method of agreement) can seldom be used as a valid 

 proof because of this difficulty of being sure that it is indeed the only 

 common factor. The snnple observation that all people suffering from 

 beriberi have diets ^\hich are low in thiamine is not proof that a defi- 

 ciency of this vitamin causes the disease, for there may be many other 

 factors in common. 



Experiments based on the method of difference provide another way 

 of elucidating cause and effect relations. If two sets of circumstances 

 differ in only one factor, and the one containing the factor leads to an 

 event and the other does not, the factor may be considered the cause 

 of the event. For example, if t^vo groups of rats are fed diets which are 

 identical except that one contains all the vitamins and the second con- 

 tains all but thiamine, and if the first group grows normally but the 

 second fails to grow and ultimately develops polyneuritis, this would be 

 a strong suggestion (but would not be acceptable as absolute proof) that 

 polyneuritis, or beriberi in rats, is caused by a deficiency of thiamine. 

 By using an inbred strain of rats that are as alike as possible in inherited 

 traits, and by using litter mates (brothers and sisters) of this strain, 



