Of the various kinds of error which arise in quantitative plankton investigations, those involved in the process of enumeration of a collection are ordinarily smallest in magnitude, though fractioning may introduce an extra sampling error, particularly when not done volumetrically.More serious are the possibilities of error in the collection of the plankters from the lake. Traps are in general more accurate than nets, but the latter are more convenient, and when made of no. 10 silk are reliable quantitative collectors of the larger organisms. The more usual no. 20 silk is very variable in efficiency. Both nets and traps appear to suffer from the ability of some plankters to see and avoid them, by day. The sampling error of a collection, considered as representing the part of the lake near which it was taken, is such as to make a single collection, containing only a moderate number of individuals, of little value in determining abundance of a species.On Cultus lake, collections taken at a single central station will indicate the average abundance of the pelagic plankton almost as precisely as would the same number of collections taken at various points throughout the whole pelagic region.