T94 J: LIONEL TAVELER [SEPTEMBER 
best chances of survival and adaptation, either to old or to new con- 
ditions, would be through conjugation. Selection would therefore 
favour conjugation (Weismann). 
3. If for some reason, possibly nutritional in origin, fission in an 
organism had not been quite complete, and the cells instead of separat- 
ing had remained together, then as each new division reached maturity 
it would divide and the process of division would continue till in- 
terfered with by some outside condition, many different forms of these 
masses of cells would thus be produced, examples of which may be 
found in the different forms of sponges. Now, if for any reason a 
curved single layer of cells was formed, it would go on growing in 
all directions until it met other cells of the same collective cell 
colony ; a multicellular growth limit would thus be reached. Now, 
assuming this growth capacity to remain constant, one of three things 
can happen. With a somewhat irregular hollow sphere of cells, it 
would be conceivable that: (1) a bending in at one of the weaker 
points, or (2) a bending out would occur, many cells being involved 
in this yielding; or (5) each cell might bud off a certain portion in- 
dependently. Of the first or outward yielding, and the formation of 
buds, we have many examples occurring in nature, as, for example, 
bud development in the hydra; of the inward yielding, the passage 
from the blastoderm to the gastrula stage, through the process of 
invagination occurring in the development of many animals, affords an 
example of the second means of satisfying this growth tendency ; 
while in the third case division of the individual cell, and separation 
from its parent tissue, occurs in the formation of red blood corpuscles 
in mammals, ete. 
4. It is obvious that the general structure of the organism would 
be least disturbed by each individual cell throwing off buds, and 
therefore the more specialised the organic structure the less likelihood 
of those organisms that reproduced by any collective alteration of the 
the organism surviving. With growing specialisation each tissue will 
become less and less able to reproduce other than its own specialisa- 
tion, hence reproduction will occur only when the buds from the 
requisite differentiations meet; now in the case of the hydra it 
appears to be only necessary to have representatives of two classes of 
cells, the ecto- and entoderm, and these thrown-off portions of cell 
structure would, when the requisite number met, owing to perhaps 
some stronger growth tendency, tend to push up the cells above them, 
and as the most likely place for the ectoderm and entoderm units to 
meet would be between these two layers, we should expect develop- 
ment to commence from this position. With increasing differentia- 
tion reproductive centres would tend more and more to be localised 
to one centre. Hence with increasing specialisation there would be 
progressively less power of local or somatic reproduction. 
5. A special kind of organism survives for two reasons: (1) 

