130 AMPHIMIXIS OR ESSENTIAL MEANING OF [XII. 



idea of a ' reducing division ' presupposes the multiplication of 

 the equivalent but individually characterized units in the germ- 

 plasm of the fertilized ^gg, and that, without this presupposition, 

 the ' reducing division ' is entirely devoid of meaning. 



If we may now feel greater certainty than ever before in 

 regarding the double division of the egg and sperm-mother-cell 

 as a ' reducing division/ we gain at the same time further proofs 

 that the germ-plasm is composed'of ancestral plasms, that is of 

 hereditary units of a higher order, each one of which, if it 

 alone dominated the ovum, would be capable of guiding the 

 whole ontogeny and of producing a complete individual of 

 the species. 



Before I attempt to show how these fundamental views throw 

 new light on the discoveries of recent years, I will say a few 

 words on the independence of the maternal and paternal 

 chromatosomes. 



According to the views which I have expressed, the nuclear 

 rods are built up of a series of ancestral plasms, which are not 

 intimately connected together, but so far as mere position is 

 concerned, are arranged next to one another. A rod does 

 not represent a kind of ' individuality ' (Boveri), by which term 

 there is certainly implied the existence gf an internal relation- 

 ship of parts to one another, according to certain laws, a 

 relationship which prevents any mechanical division of the 

 whole into equivalent parts capable of living and performing, 

 their functions. 'Individualities' as defined above are to be 

 found in my ancestral plasms, or as I propose to call them 

 shortly, the ^Ids ' ^ These cannot be divided without losing the 

 power of building up an individual, while, according to my 

 view, the series of ancestral plasms which compose the rods 

 or ^ Idmits'^ might quite conceivably be removed bodily, by 

 division occurring at any spot, and replaced by others, without 

 any loss of the essential force which controls the ontogeny 

 of the species in question. The only result of such replace- 

 ment would be to cause a more or less marked alteration in 



^ The expressions /fi? and Idant serve to recall Nageli's 'idioplasm,' 

 from which they are derived. I think it is very necessary to substitute 

 some short expressions for the clumsy ' ancestral plasms ' and ' chromato- 

 somes,' or the frequently inappropriate ' nuclear rods ' and ' nuclear 

 loops.' 



^ See the preceding note. 



