620 SUMMARY OF CURRENT RESEARCHES RELATING- TO 



which we find a small but deep pit in the ventral middle line. The 

 cells around the pit are small and spindle-shaped, or are large and 

 branched, and call to mind isolated ganglion-cells. The latter do not 

 touch the surface, but are completely shut off from it by other ecto- 

 dermal cells. All these form the first rudiment of the preoral nervous 

 apparatus of the larva. On the lower hemisphere of the larva part of 

 the ectoderm is, again, composed of as large cells as some of those in 

 the upper hemisphere, and these are found in a ventral triangular area, 

 the apex of which is turned towards the anus. This area is divided 

 by a groove into two symmetrical lateral halves. On either side there 

 is a thickened ectodermal band, the so-called ventral germ-band. From 

 this there is gradually separated off a lower layer which forms the 

 definitive mesoderm, while of the remainder the median portion gives 

 rise to the ventral nerve-cords. Before these series of differentiations 

 have been completed the ganglion-cells of the upper hemisphere have, 

 by means of their processes, become connected with one another, and 

 have thus given rise to a kind of plexus. Some, however, of the small 

 spindle-shaped cells have by their processes become closely fused with 

 the fibres of the nerve-ring, and a transverse commissure has been 

 developed by the anastomoses of the cells on either side of the hemi- 

 sphere. We have now, therefore, got a semicircular loop which 

 extends over the ventral surface of the upper hemisphere, and passes 

 by its ends into the nerve-ring. Just about this time a connection is 

 also established between the nerve-cords in the lower hemisphere and 

 the circular nerves. As this takes place at the point of formation of 

 the anterior commissure, we find that the rudiments of the cephalic 

 ganglion and of the ventral ganglionic chain are at first con- 

 nected with one another by the intermediation of the circular 

 nerves. 



Kleinenberg compares this nerve-ring in the polychastous larvae 

 with the nerve-ring of the medusas, and speaks of the upper hemi- 

 sphere of the worm-larva as the umbrella, and the lower as the sub- 

 umbrella. But if the nerve-ring is the nervous system of the larva, 

 then it has no homologue in the developed worm. " We see, therefore, 

 in the cycle of the ontogenetic development of one animal an organ of 

 the same physiological significance appearing twice over, and as being 

 formed on two different types. The larvas of Annelids possess the old 

 nervous system of the Coelenterata, while the Annelids themselves 

 have their own proper central organs, which are in no way modifica- 

 tions of the other. The organ of the lower type is developed, and is 

 functional in the larva, but in the adult it is replaced by a fresh 

 formation." 



In explanation of this remarkable modification, Kleinenberg points 

 out that variations have a definite character, which, though dependent 

 on external activities, must also be conditioned by the characters of 

 the form itself. The mere development of any new organ must be 

 accompanied by changes within, though perhaps not without, the 

 organism. There is a limit, as it were, of equilibrium to variations, 

 but this limit may be passed if the change is of advantage, and then 

 we find considerable modifications in the organism itself. The 



