34 THE PHYSICAL BASIS OF INHERITANCE 



A note is needed in regard to the misunderstanding which 

 has led many to cite cases of inheritance in unicellulars as 

 relevant to the discussion on the transmission of " acquired 

 characters." Although we can no longer say that unicellular 

 organisms are without sexual reproduction, since many exhibit 

 the liberation of special reproductive units and the occurrence 

 of amphimixis, we may still say that, apart from transitional 

 forms (like Volvox, which form colonies or " bodies " of one 

 thousand to ten thousand cells), there is among the unicellulars 

 only the beginning of the important distinction between somatic 

 or bodily and germinal or reproductive material which distin- 

 guishes multicellular organisms. This makes a notable differ- 

 ence. 



§ 4. The Hereditary Relation in the Asexual Multiplication 

 of Multicellular Organisms 



In many of the simpler, but multicellular, plants and animals, 

 a portion of the parent is separated off to form the beginning of 

 a new life. The freshwater sponge multiplies in part by minute 

 gemmules, which float away from the corpse of the parent and 

 develop into new sponges ; many polypes produce buds which 

 may be set adrift, as in the freshwater Hydra, or may remain 

 attached and help to form the great colonies that we see in 

 zoophytes and Anthozoa ; not a few worms also multiply by 

 dividing or by budding, and the examples highest in the scale 

 are found among the Tunicates, which are really vertebrate 

 animals. Moreover, in some cases where asexual multiplication 

 does not normally occur, it may still be a possibility, as is shown 

 by the fact that cut-off portions may, in appropriate conditions, 

 grow into entire individuals. Thus, two earthworms may 

 occasionally be produced by cutting one ; a sponge whicn 

 does not normally liberate buds may be cut into pieces 

 and bedded out successfully ; the arms of the starfish, whicn 



