vn ANNELIDA 135 



When we turn the flattened embryo over and view it from the 

 vegetative pole, we are able, once the 64-cell | stage has been 

 passed, to distinguish the various quadrants of the egg from one 

 another, and to tell which is A and which is B, which C and which D. 



Now the embryo has a somewhat squarish outline and the 

 rounded corners are formed by the groups of prototrochal cells 

 belonging to the first quartette. These groups are directly opposite 

 the respective macromeres from which they arose; for, if we take 

 the group derived from A for example, all its members are daughters 

 of la2 and ultimately of la. But la itself was given off dexiotropioally 

 from A; that is to say, lay above and to the right of it. The next 

 division is a laeotropic one, that is to say, as the name implies, la^, 

 the upper daughter, lies above and to the left of the lower daughter 

 lal Now this formation of the spindle with a left bend has the 

 effect of causing la^ itself to pass somewhat to the left, and thus 

 undo to a certain extent its original right-hand twist, so that it 

 eventually comes almost exactly opposite A. 



The line joining the prototrochal group of cells and the 

 macromeres constitutes a radius of the figure, and cells or cell groups 

 lying on this radius are said to be radial, and cells or cell groups 

 alternating with them are said to be inter-radial. Now it is found 

 that the third and fifth quartettes of micromeres are radial whereas 

 the second and fourth are inter-radial. The following rule then is 

 found to hold for the fate of cells forming these quartettes. In the 

 radial quartettes the cells in quadrants A and B behave alike, but 

 the cells in C and D while behaving like each other behave differently 

 from those in A and B. In the inter-radial quartettes the cells in 

 quadrants in A, B, and C, behave alike, but the cells in quadrant D 

 behave differently from each of them. 



Turning our attention now to the second quartette, and taking a 

 as example for a, b, etc., we find that 2a divides into 2a^ and 2a^, 

 one directly above the other. Each then divides laeotropically into 

 2a^^ and 2a^^, and 2a^^ and 2a^^, respectively, thus forming a lozenge- 

 shaped group of four cells. Of these four the outermost divides 

 radially into 2a^^^ and 2a^i2, lying almost side by side; whilst 2a^^ 

 and 2a2i divide obliquely into 2a^^^ and 2ai22, 2a2ii and 2a2i2 

 respectively. These cells help to form a belt of flat clear cells lying 

 just beneath the prototroch, and they divide no more till the 

 Troohophore stage is reached. The fate of the innermost cell 2a^^ 

 is different; it divides by a tangential cleavage into an outer cell 

 2a2^i and an inner cell 2a^2l Then each of these divides into a larger 

 anterior cell 2a^^i^ and 2d?'^'^^ respectively and a small posterior cell 

 2a22ip and Sa^^^P respectively. 



Exactly the same division takes place in quadrant C, and then the 

 small posterior cells sink into the blastocoele and form that part of 

 the mesectoderm or larval mesoderm, which will eventually give 

 rise to the circular and radiating muscles of the larval oesophagus. 



In quadrant B, however, 2b2^i divides into an outer and upper 



