68 MORGAN. [VoL. IX. 
single rod. The second point of difference follows as a corol- 
lary to the last statement. In Amphioxus the rods unite with 
one another at the dorsal edge of the gill-slits, forming a 
continuous series, while in Balanoglossus on account of the 
rods in the tongue-bars remaining separate a broken series of 
double inverted U-shaped structures is formed. 
A careful examination of the basket of chitin-like rods of 
Amphioxus and Balanoglossus will, I think, lead to the con- 
clusion that the structure has developed as a supporting frame- 
work for a series of gill-openings. If, on the other hand, we 
supposed a single pair of gill-openings in the ancestral form, 
and each opening supported by a double U-shaped rod, and, 
further, that the whole structure repeated itself again and again, 
still it seems highly improbable that, in the first place, the 
double rod would have appeared in each primary bar, and, in 
the second place, that the repetition could have taken place in 
the same identical way both in Amphioxus and Balanoglossus. 
It seems to me that we can draw but one conclusion from these 
data, vzz., that both forms have come frem a common progenitor 
with serial gill-slits supported by a chitin-like frame-work, and 
that this same structure in a more or less modified form is also 
inherited by the Ascidians. 
If this be granted it follows: That, since it is the last meta- 
mere of the body of Balanoglossus that carries the series of 
gill-slits and its supporting basket, this vegzon corresponds 
to a great number of metameres of Amphioxus. Therefore, 
either Balanoglossus has lost its posterior metameres, for which 
there is not the least evidence, or else the large posterior 
metamere of Balanoglossus has become split up into several 
metameres in Amphioxus. Not for a moment do I suppose 
one form to have developed into the other, but that this process 
took place in some form in the phylum that has led from the 
common ancestor up to Amphioxus. If the validity of this 
reasoning be granted, we have reached a conclusion of import- 
ance for the evolution of the problem of the metamerisation of 
Amphioxus., 
