Further Examination of the Cell-Theory 199 



half embryos and so fortli, consequently witliout tlie forma- 

 tive cooperation of other parts, so are fashioned independ- 

 ently, like the stones of a mosaic picture. Tliis takes place 

 through the 'self-differentiation' of separate cleavage cells 

 or later separate organ-foundation or in fact artificially de- 

 limited parts, and is possible only in 'typical' development 

 since 'atypical' development must proceed with far reaching 

 formative regulation." ^ And along with tliis definition of 

 mosaic work the definition of the mosaic theory should be 

 noticed. "The mosaic theory is the theory which explains 

 or at least makes intelligible the 'mosaic work.' It rests 

 upon the assumption of different qualities in tlie separate 

 cells (cell body or also cell nucleus) or organ-foundations 

 etc. capable of 'self-differentiation'." 



Obviously, were it generally true that the cells of an or- 

 ganism are as distinct as the individual stones in a mosaic, 

 not only as to their form and structure, but also as to their 

 activities, including their powers of differentiation, this fact 

 would go far toward a demonstration that cells really are 

 the "key to all ultimate biological problems" and the central 

 thesis of biological elementalism, namely, that the final ex- 

 planation of all organic phenomena lies in the elements con- 

 stituting organic bodies, would have found an almost im- 

 pregnable stronghold. On this account the mosaic theory 

 has been far more eagerly defended and combatted than its 

 merits as a strict scientific hypothesis warrant. For this 

 reason too, it will be profitable for us to devote somewhat 

 more attention to it than we could otherwise afford. 



The central facts on which the mosaic theory rests are 

 familiar to all students of embryology. Roux killed one of 

 the two cells of frog eggs when they were in the two-cell 

 stage of development, by pricking it witli a lieated needle. 

 In some cases the other cell remaining uninjured develoj)ed 

 into a half-embryo of somewliat such cliaracter as one would 

 get were he to split a noi-mally developed embryo lengthwise 



