1887] Metschnikoff on Germ-Layers. 425 
for the heaviest individuals, and consequently those laden with 
food, to move as near as possible to the centre. Another advan- 
tage would be gained if the nutritive persons were enabled to 
pursue their calling under the best possible circumstances. Now 
we know that many Flagellata pass from a monad into an amce- 
boid stage, and it is especially in the latter that they feed. I 
will quote Biitschli’s account (23): “In a large number of mo- 
nads food is only taken in at a particular spot, which is almost 
always at the base of the flagellum. There is no doubt of this 
fact, and it is equally true, on the other hand, that certain of 
these forms, which have been observed to pass occasionally into 
the amceboid condition, are, during this condition, able to take 
in food at other points of the body.” It is easily conceivable 
also that the nutritive individuals should multiply more rapidly 
than the others, and a connection might thus be established 
between an unusually large food-supply and multiplication. Still 
another cause for the occurrence of immigration is found when 
we consider that a colony could not increase its superficial area 
beyond certain limits. This fact must have exerted great influ- 
ence on swimming colonies especially. Since increase in the 
number of individuals in our colony was sure to lead to increased 
activity, it was manifestly advantageous that cells, for which no 
room could be found at the surface, should adapt themselves to 
a life in the interior. The difference between individuals of the 
same stock was probably for a long time a mere quantitative 
difference: the locomotor cells attracted particles of food by 
means of the currents set up by their flagella, and also took in 
some of the smaller particles, as in certain Ccelenterates to-day, 
where the ectoderm cells occasionally take in bits of food (9). 
The ameeboid individuals inside were, on the other hand, able to 
devour larger bodies, which the locomotor cells could not mas- 
ter. When thus engaged, the amceboid cells were very probably 
near the periphery, and no doubt made use of the numerous 
` pores penetrating the superficial layer in order to get at the bits 
of food lying on the surface. For analogies to such a perform- 
ance I may refer to the fact that in Halisarca pontica, under 
certain circumstances (8), foreign bodies are devoured by amee- 
boid cells, and that in many of the lower animals the same is 
done by mesoderm cells, The differentiation gradually made 
greater progress in the path thus marked out. The locomotor 
VOL. XXI,—NO. 5. 29 
