436 University of California Pivblications in Zoology [Vol. 18 



plates seems to increase most rapidly, and the latest additions to this 

 row seem to be produced by the addition of plates, small in size at 

 first, on the ventral side at either end of the precingular row and next 

 to the rhomboid plate, which must be considered as belonging to the 

 apical row but which interrupts the precingular row on account of its 

 extent from the apex to the girdle. 



At a certain stage in the progressive multiplication of the plates 

 of this shell in this family, an accessory plate or an incomplete row 

 of plates makes its appearance either on the dorsal side of the hypo- 

 theca between the antapical and postcingular rows or more often on 

 the dorsal side of the epitheca between the apical and precingular 

 rows. The number of plates in the postcingular accessory row rarely 

 exceeds two in number and that in the anterior accessory row rarely 

 exceeds three in number. 



The genus, Periddnium, by definition, perhaps largely fortuitous 

 or conventional in its formation, is made to include forms having a 

 definite number of plates in the four main rows, no posterior acces- 

 sory plate and either two or three anterior accessory plates. In the 

 greater number of known species coming under this general defini- 

 tion there are three anterior accessorj^ plates arranged continuously 

 in a row over the dorsal part of the epitheca between the apical and 

 precingular rows. Those species in which the number of accessory 

 plates is less than three are readily related to the major part of the 

 group. This reduced number of accessory plates is to be regarded 

 as a preliminary condition introductory to the later condition of three 

 such plates represented in the greater portion of the genus, or as of 

 secondary acquisition due perhaps to more or less of a readjustment 

 upon transfer to fresh-water conditions from estuarine conditions on 

 the part of fresh-water forms, most of which present such a reduced 

 number of accessory plates. For some reason the number of these 

 accessory plates seems in the majority of the species of this genus to 

 have become three in number but never to have exceeded this number. 

 No instance is known to the writer of only one accessory plate, but 

 at least one form has been reported with no accessory plates, which 

 nevertheless conforms to the other characters of the definition of this 

 genus. 



