Differentiation of Neuroblasts 55 
is naturally led to inquire if there is any possibility of placing a 
different interpretation on these observations. 
Harrison’s experiments, described above, in which portions of 
the medullary tube placed in a drop of lymph developed nerve 
fibers, are open to the objection that the lymph necessarily con- 
tained products of the metabolism of various organs of the body, 
and it is therefore not certain, indeed it is improbable, that the 
neuroblasts were removed from the influence of end organs whose 
physiological activities were similar to those which they normally 
innervated. Such an experiment is, therefore, not crucial, though 
I believe it furnishes the strongest evidence in favor of self- 
differentiation that has yet been given. ‘The experiments in which 
pieces of the medullary tube were transplanted to abnormal posi- 
tions inside the body of the tadpole are open to the same objection 
to a still greater degree. 
A number of points require further investigation before Braus’s 
experiments in removing the limb of Bombinator can be regarded 
as proving conclusively that the neuroblasts are self-differentiating. 
His conclusions rest mainly on the fact that ten days after the 
removal of the limb the brachial plexus is as large on the operated 
as on the unoperated side of the animal, although some weeks 
later, after metamorphosis, it is smaller. “Die Verschmachtigung 
bei dem 4lteren Objekt,”’ he says, “ist demnach eine sekundare.” 
But before this can be accepted the degree of differentiation of the 
nervous system ten days after the operation must be considered, 
and the impossibility of stimulus from any brachial muscle or 
other muscles be established. 
In regard to the specimens examined near the end of metamor- 
phosis he says, “Derselbe (the brachial plexus) ist allerdings 
dinner als derjenige der normalen Seite der Larve und auch 
ventralwarts mit Muskelanlagen der Bauchwand in Verbindung.” 
In the specimen killed 10 days after the removal of the limb he 
states that the trapezius and interscapularis muscles are present, 
so that in neither of these cases are the neuroblasts removed from 
all muscular influence, and experiments on the chick have shown 
that the defect in the size of the nerves is not proportional to the 
defect in the musculature. A certain degree of development may 
