200 Edmund B. JVilson. 



meres after each division, transfering them to normal water at the stage de- 

 sired and all my critical results have been thus attained. The mortality is very 

 large, since the blastomeres seem to suffer severely in the change from the ar- 

 tificial to the normal water, and is greatest in cells from the vegetable hemi- 

 sphere; hence my failure thus far to isolate successfully the second somatoblast 

 (or primary mesoblast-cell, 4d), in some respects the most interesting of all 

 the cells. For the latest stages I did not endeavor to isolate the cells at all, 

 but allowed the eggs to develop for 24 hours in the artificial water, from time 

 to time separating the cells by jets from a fine pipette. 



Most of the studies on isolated blastomeres were made on Patella, since 

 with Dentalimn most of my time was given to experiments on egg-fragments. 

 For preparation of the Patella eggs I found no better method than the simple 

 one employed by Patten ('85) ot acetic acid and glycerine. The eggs were 

 placed in a watch-glass nearly filled with sea-water, two to four drops of glacial 

 acetic acid added, followed by successive additions of dilute glycerine gradu- 

 ally replaced with strong glycerine. This renders the embryos perfectly trans- 

 parent, with sharply marked cell-boundaries, and often gives preparations of 

 admirable clearness. A slight stain with acetic carmine often adds consider- 

 ably to the effectiveness of the preparation for a time, though they subse- 

 quently deteriorate, and for most purposes the stain is superfluous. 



II. 



PRELIMINARY NOTES ON THE NORMAL DEVELOPMENT. 



Unfortunately the cell-lineage of neither Patella nor Dental- 

 iiim has been worked out. Patten's early paper on the embry- 

 ology of Patella ('85), excellent as it is in many respects, leaves 

 this part of the development nearly untouched, and the same is 

 true of the still earlier paper of Lacaze-Duthiers ('57) and that 

 of Kowalewsky ('83) on Dentalium. Everyone familiar with 

 work of this type will appreciate the fact that to work out the 

 cell-lineage fully would require prolonged study, and both the 

 forms here dealt with present peculiar difficulties in the later 

 stages. The time at my disposal has only allowed me to deter- 

 mine the main outlines of the cell-lineage, including details es- 

 sential to the interpretation of the more important experimental 

 results. Fortunately, however, Robert ('02) has recently pub- 

 lished a detailed study of the cell-lineage of TrocJiuSj which agrees 

 so closely with that of Patella that it may be taken as a standard 



