MENDEL'S LIFE AND WORK 21 



pairs of contrasted characters, without being dis- 

 tracted by any other changes which may result as 

 the effect of the cross. This may seem a very 

 obvious method of procedure — something, in fact, 

 very like taking the old advice of breaking the 

 sticks one by one instead of trying to break the 

 whole bundle at once — but in reality it involves an 

 assumption which is even yet regarded by many as 

 somewhat of a heresy, namely, that an organism 

 can be regarded in any degree as a bundle of separable 

 unit characters. Luckily^ in the case of Mendel's 

 peas, the selected characters proved to be simple and 

 completely separable, and so the problem of their 

 transmission was reduced to its simplest terms. 

 Mendel himself, however, seems to have realised 

 that it might be very difficult, in certain cases, to say 

 offhand what were the characters which were treated 

 in heredity as units, for he says " the uniformity of 

 behaviour shown by the whole of the characters 

 submitted to experiment permits, and fully justifies, 

 the acceptance of the principle that a similar relation 

 exists in the other characters which appear less 

 sharply defined in plants, and therefore could not be 

 included in the separate experiments." Later experi- 

 ments have abundantly shown that it is practically 

 impossible to say at first whether any particular 

 character that can be visually recognised is or is not 

 a unit character, and it therefore becomes an essential 

 part of any investigation on heredity to ascertain 

 what are the fundamental characters involved. 



