LINEAR AND SUPERFICIAL AGGREGATES. 1 9 



tion, a roiv of four cells will result; the two inner cells 

 would be disks or short cylinders, while the two outer would 

 be hemispheres (fig. 23, B). (b) But if (as is actually the 

 case in Pleurococcus, B, fig. 18) the new partitions are at 

 right angles with the first, the result is a cluster of four cells, 

 each of which is a quarter of a sphere (fig. 23, C). 



Should a third division occur, it is conceivable that the 

 new septa might be placed parallel to those already formed, 

 in case a ; or parallel to one set and at right angles with 

 the other, in case b ; or at right angles to both, in case c. 

 In the first instance there would be formed a row, or filament, 

 of eight cells; in the second, a sheet of eight cells; or, in the 

 third, a mass of eight cells. This exhausts the possibilities in 

 the position of successive partitions. If other divisions 

 occur, they will necessarily be more or less nearly parallel to 

 some one of the first three sets.* 



The structures resulting from cell-division where the cells 

 remain united are conveniently designated as follows: (1) 

 cell-rows, filaments, or linear aggregates, arising by division 

 in one plane; (2) cell-surfaces, or superficial aggregates, 

 arising by division in two planes; (3) cell-masses, or solid 

 aggregates, arising by division in three planes. 



It is manifest that there are likely to be all degrees of union 

 remaining between the cells of linear and superficial aggre- 

 gates, and that the extent and firmness of such union will 

 depend largely upon the character of the wall. As in every 

 other case, the artificial distinction between cell-colonies and 

 cell -aggregates is bridged by all manner of intermediate 

 forms. 



Filamentous algae. 



There is a large number of plants in which the vegetative 

 body throughout life has the form of a filament. The green 

 *The formation of partitions at angles other than 90 or l8o° to pre- 

 ceding ones would not affect the genera] result, luu would only render 

 the form of the product, as well as of the individual cells, less regular. 



