APPENDIX F. 



PHYSIOLOGICAL (OR CONSTITUTIONAL) UNITS. 



THERE has recently come before me a fact which has a signifi 

 cant bearing on the hypothesis of Constitutional units : serving, 

 indeed, to give an apparently conclusive proof of its truth. Before 

 stating it, however, I may with advantage re-state the several 

 evidences already assigned in support of it. 



1. First comes the a priori reason. These units in the germ of 

 an organism which cause development into a special structure, 

 cannot be chemical units cannot be simply molecules of proteid 

 substance in one or other of its forms ; since these are not special 

 to any type of creature but common to all creatures. Nor can 

 they be what we may call morphological units the cells or proto 

 plasts ; because in the early stages of development the cells of 

 one organism are indistinguishable from those of others, and 

 because were cells the units of composition there could be no 

 interpretation of what are called unicellular organisms nothing 

 to account for the innumerable varieties of them. Hence, of 

 necessity, the structural elements of which each organism is built, 

 being neither proteid molecules nor cells, must be something 

 between them : probably some complex combination of different 

 i&omeric forms of proteids. 



2. That units of such natures are the essential components of 

 each species of organism, is shown by the fact that in low types of 

 creatures, little differentiated into special tissues, any considerable 

 portion of the body will, when separated, beofin to assume the 

 structure proper to the species a truth recently shown afresh by 

 Prof. T. H. Morgan s experiments on the regeneration of Planaria 

 maculata (already referred to in 206) showing that various frag 

 ments cut out develop into new individuals, and that when, being 

 too small they die before doing this, there is always an abortive 

 attempt to assume the specific structure. 



3. This truth that a portion of undifferentiated tissue, if ade 

 quate in quantity, assumes the structure of the type, illustrating 



612 



