802 



THE AMERICAN NATURALIST [Vol. XL 



In typical cases the new column of plates is initiated by an area 

 liaving but five instead of* six sides. To compensate for this 

 mechanically, the preceding area has seven sides (Fig. 2, a). 

 This equating method of heptagons and pentagons is typical 

 whenever a single column is added in one spot. In certain cases, 

 however, more than one column may originate, even with the same 

 area, and then various irregularities are taken on (Fig, 2, h). In 

 such cases, ;irc;i> with hut four sides are met with occasionally 

 instead of the iiotmal pciitagoual areas. Besides this variation 

 in the number of sides of the areas in different portions there seems 

 to be a definite alternation of the columns, to the right and left of 

 the axis in which they are added. This applies of course only to 

 those columns added successively in one of the three directions 



Altogether, the test of Arcella is far from the siinj)le hexagonal 

 structure figured by Leidy and other authors. Its complexities 

 are worthy of further study and comparison with the tests of other 

 rhizopods. 



