Manchester Memoirs, Vol. I. (1906), No. ||. 19 



knowledge we have of the causation of natural phenomena 

 and so to obtain what control we have of the operation 

 of natural processes. Just because we know that ex- 

 planation is after all only description, it does not follow 

 that we should abandon the attempt to account for things. 

 We are now in a position to classify Laws of heredity 

 under these two headings. 



Phvsiolog ical. Statistical. 



(a) Mendel's Law. (a) Galton's Law. 



(/;) The Law of Contribution. {i>) Pearson's Law. 



Now inasmuch as of physiological Laws d has been 

 shewn to be invalid, and of statistical ones a has been 

 shewn to be less comprehensive than d, the discussion of 

 the mutual relation of physiological to statistical Lav^s of 

 heredity resolves itself into a discussion of the relation of 

 Mendel's Law to Pearson's Law of Ancestral Inheritance. 



4 (/;). Mendel's Law True of Units : Pearson's of 

 Masses. 



Discussing a little while ago with a friend the relation 

 of Mendelism to biometry, I suggested as the briefest 

 possible statement of the difference between the two that 

 Mendelism treated of units, and biometry of masses of 

 units : my friend replied. ' You speak of the animals and 

 plants with which you deal as breeding true to such and 

 such a character. What do you mean by this statement? 

 Are the offspring absolutely identical with their parents 

 in respect of the character under consideration? If they 

 are not, how like are they ? and how is the degree of this 

 similarity to be measured, except by biometric methods?' 



I saw that there was truth in what he said ; but I 

 could not see the relation which the idea in his mind 

 bore to my original idea, to which I still adhered, 



