578 DR. J. B. HICKS ON THE GONIDIA AND 



each patch, a large stream of these Chlorococcus-like bodies may be seen, running 

 downwards. I have not separately drawn these cells, because they are so like those on 

 PL LVIII. fig. 18 c, to which I refer the reader. 



Sometimes on trees I have observed that, instead of producing the globular free cells 

 just described, as occurs in most situations, the whole cells of the filament, which is 

 generally tapering towards the extremity, undergo this quaternary form of segmenta- 

 tion in situ. It will be first noticed that the cell-walls become thicker, the contents less 

 granular, while the whole filament increases in diameter, and that then the contents are 

 segmenting (PL LVIII. fig. 18 a). After a time the parent-cell waUs dissolve away, 

 and the subdivisions become free (PL LVIII. fig. 18 b). From this point the description 

 applied to those cells which were free before division corresponds so exactly as not 

 to need repetition (PL LVIII. fig. 18 c). By this means, as by the other, large por- 

 tions of bark of trees are covered with the Chlorococcus-like bodies, which multiply also 

 indefinitely. 



But in some filaments there is a still more unsuspected change, namely, in the pro- 

 duction of cells of Glceocapsa. The segmentation proceeds within the filament, as in the 

 instance just quoted ; but the divisions become invested in a gelatinous envelope, while 

 the parent-ceU wall breaks up. These Gloeocapsa-like bodies then become free, and 

 continue the segmenting process as in Gloeocapsa. This I have shown on PL LVIII. 

 fig. 19 a. It is a condition by no means rare in the winter months : considerable masses 

 of these bodies are to be found so produced. 



I have frequently seen Glceocapsa polydermatica (Kiitzing), &c., formed, as weU. as 

 other so-called species. After frequent segmentation, the cells are imbedded in an 

 indefinite mass of gelatinous substance. 



But there is a variety of this " Gloeocaptsa-iovm&iion " frequently met with (PL LVIII. 

 i\g.20dcl). 



The cells of a filament in one or in every part at once begin the process of quaternary 

 segmentation, as before noticed, at first regularly, but shortly after irregularly ; besides 

 this, a certain amount of free-cell formation goes on within the divisions (mother cells) ; 

 so that it is difficult to say which kind of cell-formation predominates (PL LVIII. 

 tig. 20 a a). In this manner large irregular masses of segmentary cells are produced, 

 like some of those resulting from segmentation of the so-called Palmellacese (PL LVIII. 

 fig. 20 5 6). The cells set free from them are either Chlot^ococctis-like cells of variable 

 size (PL LVIII. fig. 20 cc), or they are like Glceocapsa, undergoing segmentation in their 

 variable manner (PL LVIII. fig. 20 dd). 



These changes can be readily observed in the colder months. They frequently, by 

 distortions in all directions, produce a mass whose origin might be very doubtful to 

 determine, were it not generally possible to find some small part retaining the original 

 filamentous condition (PL LVIII. fig. 20 a a). 



Up to this point of the observations I have made, with the exception of that just 

 noticed, the efforts made by the confervoid filaments towards multiplication have been 

 carried out upon the plan of the " parietal cell-formation " of Nageli, and that principally 

 upon the quaternary mode. 



