220 
R. Ruggles Gates . 
forth above, implying that germinal changes arise in a fixed life- 
cycle through a change in the cell unit which is represented in every 
nucleus and therefore may modify every stage of the new organism. 1 
On the other hand, if the more fundamental facts of recapitulation 
mean anything, they imply that at some time an actual lengthening 
of the life-cycle has taken place, either by the addition of cell 
divisions at its end or by their intercalation at some point. Such 
a process can not easily be accomplished by a variation in the 
structure of the cell or nuclear unit itself, but must rather be the 
result of the organism as it were overcoming its cell shackles and 
by its own energy producing new developments, though such novel 
additions are themselves cellular in structure. 
From the time the foundations of the cell theory were laid by 
Schleiden and Schwann in 1838-39, its universal sway was scarcely 
questioned for over half a century. During this period it was 
established that cells arise only from the division of previous cells, 
and the cell theory culminated in such conceptions as the physio¬ 
logical division of labour among cells, the mosaic theory of 
embryonic development, the individual as the sum of the activities 
of its various cells. In short, it came to be assumed that cells 
make the organism, while the contrary fact that the organism after 
all makes its cells was tacitly or explicitly denied. In the same 
way it was assumed that each cell of a multicellular organism 
necessarily corresponded with the whole organism in the Protozoa. 
One of the first reactions from this extreme development of the 
cell theory, which made the organism not a master in its own house 
but a slave of its constituent cells, was a well known paper by 
Sedgwick (1894) on the inadequacy of the cell theory of development. 
His views were based on studies of the embryos of Peripatus 
capensis , which he believed were essentially coenocytic in structure; 
also upon the development of mesenchyme and nervous tissue 
in Elasmobranch embryos. As far as Peripatus is concerned, 
delicate cell walls have recently been demonstrated (Glen, 1918) 
both in the ectoderm and endoderm layers, by careful preparations 
and the use of an immersion lens. But their demonstration does 
not affect the fundamental question of the relation of the cell to 
the organism, for, to mention only a few cases, it is well known 
that in the formation of the blastoderm of the insect egg, as well 
as in the early stages of development of the female gametophyte 
and proembryo in Gymnosperms, a stage is passed through in which 
1 It appears that, especially in animals, the change is frequently visible 
or obvious only in certain organs. 
