6 PROCEEDINGS OF THE NATIONAL MUSEUM V0U78 



tunately, the mandibles can not be seen in many species known only 

 from a single poorly preserved type specimen so the value of this 

 character can not be fully utilized. 



Antennae. — These have been used to a great extent in this paper, 

 for with few exceptions the antennae of all the type specimens can 

 be more or less clearly seen. However, in the use of the illustrations 

 and of the table of measurements, some allowance should be made, 

 since many of the drawings were made from poorly mounted speci- 

 mens. Even when actually viewing the specimens, it is not always 

 possible to perceive the true proportions since there may be a fore- 

 shortening due to an oblique position, or the lateral aspect showing 

 the greatest width may be seen in one specimen and not in another. 

 In a few cases some of the antennal joints were obscured by the 

 head or other parts, and such obscured parts are indicated in the 

 drawings by broken lines; such reconstruction is more or less hypo- 

 thetical and can not always be interpreted as being accurate. Also, 

 some of the antennae had shriveled before being mounted in balsam 

 and in such cases the outlines were drawn with the idea of depicting 

 the parts as they appeared before shrinkage. In preparing the 

 descriptions of the antennae, a camera lucida drawing was first made, 

 then the individual joints were measured; the descriptions were based 

 on the drawings and measurements thus obtained. As a result, 

 differences were noted that were not appreciable to the unaided eye. 

 The detailed measurements were obtained for the purpose of making 

 the descriptions more accurate and not with the intention of includ- 

 ing them in the paper for publication. It has been found that these 

 measurements are of some aid in separating certain closely allied 

 species and that the descriptions can be better understood if the 

 actual dimensions are given, since very slight differences, not always 

 ajiparent to the unaided eye, have been described. The species of 

 Coccophagus may be divided into two groups on the basis of anten- 

 nal characters. Many of the small yellow species have the pedicel 

 longer than the first funicle joint, the first funicle joint the shortest, 

 the second and third funicle joints successively longer. The great 

 majority of species have the pedicel distinctly shorter than the first 

 funicle joint, the first funicle joint the longest, the second and third 

 funicle joints successively shorter and wider. A few species are 

 intermediate and not easily assigned to one group or the other. 

 With few exceptions, the funicle joints are plainly longer than wide. 

 The number and arrangement of the sensoria is of taxonomic value 

 and this detail is shown in the drawings. 



Palpi. — Maxillary palpi two-jointed, labial palpi one-jointed. The 

 relative lengths of the palpal joints differ among certain species but 

 this character has not been used as the differences are not especially 

 marked and frequently none of the specimens are mounted in a way 

 that reveals the mouth parts. 



