SHELL-GLAND OF CYCLAS AND PLANULA OF LIMNJEUS. 325 



canals (the urnieren) were not present. The fact that they 

 are figured by Mr. Rabl has led me to re-examine these 

 stages in specimens obtained from the same locality as those 

 which I previously studied. 



I find that the ' nephridia' or primitive kidneys, known 

 as Stirbel's canals, are present in the position described by 

 Rabl. They escaped my attention owing to the fact that I 

 was engaged with other points, viz. the shell-gland, intestine 

 and velum. The otocysts and pedal ganglia do not appear 

 until later. 



It is a pleasure to me to have finally to draw attention to 

 the recent publications of Professor Fleming, ' Wiener 

 Sitzungsber.,' Bd. 71, 1875, and ' Zeitschr. fur Wiss. Zool.,' 

 1875) relative to Molluscan embryology — since he has been 

 led to accept the accuracy of my observations and finds in 

 them the explanation of some difficulties connected with the 

 interpretation of his own results obtained in the study of 

 Unio and Anodon. 



Formation of the Gastrula in Limnceus, — In my paper on 

 the embryology of Limnseus I pointed out the difficulty 

 which I had experienced in tracing the precise mode of for- 

 mation of the invaginate planula or gastrula from the morula 

 of that animal. 



I have endeavoured to overcome that difficulty since, but 

 cannot consider what I now have to submit as final. 



The interpi'etation which I have given of figs. 4, 5, 6, pi. 

 xvi, of my memoir in vol. xiv of this Journal, is erroneous, 

 and to a certain extent the drawings are incorrect. Fig. 4 of 

 that plate is more correctly represented in fig. 6 of the pre- 

 sent paper — and fig. 5 of that plate by fig. 5 of the present. 



I am not able to agree with those who maintain that there 

 is a distinct blastula phase in the development of Limnaeus 

 in which the cells are equi-formal. I find that continuously 

 the cells of the antiklastic pole (that is to say those opposite 

 to the pole at which the directive corpuscles escape) exhibit 

 a larger size and a greater abundance of brown granular 

 food-material than do those of the klastic pole. 



After the stage represented by fig. 4 of the present plate, 

 the cells of the klastic hemisphere commence to flatten out 

 and to recede from those of the antiklastic hemisphere, leav- 

 ing a central space between the two groups — which is filled 

 with liquid and is traversed by tapering processes of the 

 attenuated cells of the klastic pole. These no doubt are 

 the first indications of the filaments connecting endoderm 

 and ectoderm which are figured in pi. xvi, fig. 14, and pi. 

 xvii, fig. 19, of my former paper. The klastic hemisphere 



