246 ESSENTIALS OF BIOLOGY 



so found were originally mutants. They were bred with brown- 

 eyed amphipods and the red eyes reappeared in the progenies 

 and the progenies of the progenies, in the Mendelian manner, 

 so that, after the original appearance of the mutation red-eyed 

 Gammari were regarded as Mendelian variants. This notion of 

 real novelties of structure, or of the appearances of mutants, is 

 essential to any working hypothesis of transformism that we can 

 make. Later we shall consider the conception in greater detail. 



Random fliictiiants . The mere form of a frequency distribution, 

 such as those that we have studied above, suggests the notion of 

 random variation. We can best illustrate this notion by regarding 

 the results of operation of some process, or mechanism, that is 

 designed to produce things that are intended to be replicas of 

 something, say, a minting machine that produces coins that are 

 expected to 'be of the same weight and form. Such similarity 

 of product cannot be attained and if a number of the coins are 

 precisely measured it is always found that the individuals fluctuate 

 in weight, etc., round about some mean values of the characteristic 

 that is measured. There will be a small range of values, — e to 

 mean, and mean to + ^ (see Fig. 34, 2). The number of indivi- 

 duals that display this small range of values of the characteristic 

 will be greater than the number of individuals that display any 

 other equally small range of values. The number of individuals 

 displaying any similar range of values is less the further removed 

 is that range from the mean, or central range. Such a frequency 

 distribution is easily described mathematically as due to the results 

 of operation of a great number of small independent causes and 

 most frequency distributions that represent the organic variability, 

 in respect of some structural character — that is, the results of the 

 measurements of some character in a large number of individuals, 

 taken at random from a racial population, display this form of 

 distribution. Most of the individual organisms included in such 

 a distribution will be random fluctuants. It is very probable 

 that any such fluctuant (or random variant, or variant by random 

 fluctuation) will not have progeny, or progeny of progeny, that 

 display the variation in the same degree. In other words, the 

 random fluctuations are not inherited. This is a result of experi- 

 ence. The variations of numbers of vertebrae in, say, Greenland 

 cod are random fluctuations. 



Fluctuants by acquirement. Lastly, we consider those organic 



