GENUS CODOSIGA. 337 



forms in which the zooids are ahnost, but not quite, sessilely attached to the summit 

 of the primary stalk, as in C. Iwfiytis, to those in which the latter structure is con- 

 spicuously branched, as in C. umbellata, and as shown by such types as C.pyrifonnis, 

 C. grossulariata, and C. catidelabi-um, hitherto unknown to Stein, is however so 

 gradual that an independent generic title in the present instance cannot be con- 

 sistently maintained. 



In examples of Codosiga uvibellata examined by the author while going to press 

 —November 1880, received from Mr. John Hood, of Dundee— a phenomenon has 

 been observed not previously recorded of any other representative of the present 

 Flagellate order. These having been submitted to somewhat undue pressure, threw 

 out around their bodies a hyaline film of sarcode, which imparted to them the aspect 

 of being enclosed within independent sheaths or lories, ultimately retracting both their 

 collars and flagella. This pressure being removed, the bodies resumed their accus- 

 tomed shape and the flagella and collars were again extended. In the delineations 

 of Codosiga botrytis given by O. Biitschli, one abnormal example, figured and 

 described as possessing a delicate viscid case, apparently represents a closely parallel 

 condition of metamorphosis. 



Codosiga allioides, S. K. Pl. II. Figs, i and 2. 



Animalcules as in C. botrytis, but associated upon a multicapitate pedicle 

 or zoodendrium, the main stem developing from one point as many as ten 

 secondary branches of equal length, at the extremities of which the animal- 

 cules are grouped in subsessile social clusters; contractile vesicles and 

 endoplast conspicuous. Length of bodies 1-1650". Secondary stalks 

 six or seven times, and primary stalk over twelve times the length of the 

 supported zooids. Hab. — Fresh water. 



The umbellate zoodendrium of this variety exhibits a plan of ramification that 

 corresponds substantially with the floral umbel of the genus Alliiiin and its allies 

 among vegetable types, and upon which account the present specific title has been 

 adopted. In another direction this species may be said to present the appearance of 

 a number of colony-stocks of Codosiga botrytis, united at the bases of their respective 

 pedicles to one common main rachis. In the single example that has yet been met 

 with, there were no less than ten of these branches bearing each from three to six or 

 seven animalcules. The length of each secondary stalk was rather longer in pro- 

 portion than the single one of C. botrytis, while the main rachis measured a little over 

 twice the length of the secondary ones. A corresponding diversity in size also subsists 

 between the individual zooids of this type and those of C. botrytis, the latter being 

 considerably smaller. The single specimen here figured and described was found 

 growing on Nitclla taken from a pond in the neighbourhood of London, supplied to 

 the author by Mr. William Gay, F.R.M.S. From Codosiga umbellata, with the 

 irregular growth-form of which, as reproduced from Stein's work at PI. IV. Fig. 5, 

 it to some extent agrees, the present species may be readily distinguished by the 

 lax and undulating instead of rigid and rectilinear character of the supporting 

 pedicle. 



Codosiga cymosa, S. K. Pl. III. Figs. 3-7. 



Zooids symmetrically ovate, stationed separately, upon short inde- 

 pendent footstalks, at the extremities of a cymose or corymbiform, pro- 

 fusely branching pediclcor zoodendrium. Length of bodies 1-5000", of main 

 rachis 1-500" to 1-250". Hab. — Salt water. 



The number of animalcules included in a single colony-stock of Codosiga cymosa 

 exceeds that of any other species of the genus yet discovered, the luxuriantly branch- 

 ing pedicle not unfrequex.tly supporting, as shown in the accompanying plate, as 



Z 



