1917] Taxonomic Value of Antennal Segments 267 



wild grape, about ten miles south of Columbia, Missouri, and 

 the others in and around Columbia, Missouri, during mid- 

 summer upon Lactuca canadensis, L. sagetifolia, Geum can- 

 adense and Solidago stricta. Careful mounting and examination 

 of the materials on the above mentioned annuals and on the 

 perennial Solidago stricta readily disclosed the fact that these 

 four lots belonged to the same species, Pseudococcus shaferi 

 species novo, and the antennal graphs further substantiated 

 this result beyond a doubt. The material from wild grape was 

 also mounted, there being one adult female and several nymphs. 

 The antennal graph, as first made, was misleading, due to the 

 fact that the antennal measurements of the adult and of the 

 young were plotted together, thus hiding the true antennal 

 graph of the adult. So this coccid on wild grape was at first 

 referred by the writer in his manuscript to a new species. It 

 was only after replotting the antennal measurements of the 

 adult's segments that the similarity to the graph of Pseudococcus 

 shaferi species novo was observed. Careful comparisons of the 

 descriptions also showed identical details, such as the alveolated 

 posterior coxae, the size of the leg segments, the anal ring and 

 anal lobe setae, the conical spines in the cerarii, the spiracles, 

 the obscureness of the cerarii anterior to those on the anal 

 lobes, etc., so that undoubtedly the coccid from w^ild grape is 

 also Pseudococcus shaferi species novo. The reason for the 

 apparent difference in food habit is that this coccid, feeding 

 upon annuals and perennials throughout the growing season, 

 forsakes such situations late in the fall and ascends to places 

 which will insure dryness and protection throughout the winter 

 and early spring and here forms its ovisac, the spring brood 

 descending to feed upon the preferred food plants, which are 

 either annual or perennial. Such procedure has been observed 

 in the case of Trio?iymus americanus (Ckll.) which left its food 

 plants — several kinds of grasses — and^ ascended the trunks of 

 different species of trees late in the fall, and upon which their 

 ovisacs were formed. 



Other similar instances might also be cited, but the writer 

 feels sure that the value of the graphic representations of the 

 lengths of a'ntennal segments has been clearly shown in the above 

 cited case. 



It will be noticed in the discussion of the accompanying 

 graphs that the writer gives the antennal formula of the mean 



