250 NATURAL SCIENCE [October 
(7) That when not required in cell development they will tend 
to pass into the system of the organism, and that when 
suitable conditions arise they will tend to reproduce the 
cells from which they are derived. 
In the earlier forms of life these units will diffuse themselves 
throughout the organism (Protozoa), but as differentiation occurs 
these units will tend to become localised at one or more places 
(Hydrozoa). Of these places one will become more important 
either from habit or position, and this will become fixed and 
subsequently specialised (ovary or testis). The cell differentiation 
will at last become so great that it will stop all reproduction of 
parts except at the specialised centre. Partial renewal of limbs, etc., 
in the earlier vertebrates becoming rarer and ceasing altogether as 
we ascend to the higher vertebrates. 
(8) That these reproductive units having once started a 
phenomenon in any given direction, the direction will tend 
to be kept up and continued by physiological laws. 
(9) That each unit would tend from habit to occupy in a new 
organism a position similar to that which it occupied in 
the parent. 
This theory would explain the constancy of type, as there 
would be a continually increasing balance in favour of heredity. It 
would satisfactorily explain the recapitulation theory of embryology. 
It would account for the recognised antagonism existing in both 
plants and animals between the reproductive and bodily growth, and 
it would afford an explanation of growth in abnormal situations. 
In conclusion, I think it will be found that we are brought 
back to a closer study of the causes of variations as the only 
satisfactory means of solving the fundamental problem of use- 
inheritance. J. LIONEL TAYLER. 
