6 BULLETIN fi3, T^NTTED STATES NATIONAL MUSEUM. 



Dyke, Charles Fiiclis, and the hite Beverh' Letcher for their en- 

 couragement and advice. Without this material and moral aid the 

 present work could not possibly have been as complete as it is. 



METHODS OF STUDY. 



All comparative measurements have been made from outlines 

 drawn by aid of a binocular microscope with either a ^, 1, 2, or 3 

 inch objective and a camera lucida. The le^s were measured directly 

 with a finely divided metric ruler, under a hand lense. 



Fresh specimens were hardened in alcohol or a two per cent solu- 

 tion of formalin; dried specimens were (juickly relaxed in hot water. 



To determine the relative length of the antennte — the parts being 

 relaxed — the head is either retracted or extended so that the apical 

 margin of the pronotum corresponds to the j^osterior margin of the 

 tempora. (See line C, Plate 8, fig. 1.) The antenna is then 

 thrown back over the disc of the pronotum so that it falls a little 

 within the basal angles. If it does not reach beyond the apical three- 

 fourths of the pronotal disc, it is said to be short ; if it reaches to the 

 basal margin it is moderate in length ; if beyond the basal margin it 

 is long. Qualifying terms are used to express nearness to these 

 points. 



This method is arbitrary, but sufficiently correct for all practical 

 purposes, if care be taken to have the head in the proper position. 



In determining the relative width and length of the head a tracing 

 is made, care being taken to have (he surface plane parallel to the 

 focal plane of the microscope. The drawing is then treated as in 

 fig. 1, Plate 8. A line is drawn transversely, just touching the pos- 

 terior margins of the eyes and called tlie post-marginal ocular line, 

 fig. 1Z>, Plate 8. A second line is drawn from a point on the post- 

 marginal line midway between the eyes to the middle of the anterior 

 margin of the epistoma, and termed the interlocular line (fig. !</, 

 Plate 8). Line h practically divides the head into e(|ual halves, 

 near enough for our purpose. If the distance across the widest or 

 most prominent part of the eyes is tAvice the length of line (7, the head 

 is said to be moderate in width; if less, the head is long^ and if 

 greater it is wide or broad. The pro])er qualifying terms being used 

 to express nearness to or extremes of these relative dimensions. For 

 example, the head may be said to be longer than wide, somewhat 

 longer, or distinctly longer than Avide. Tlie head is sub(|uadrate, 

 and if line a be one-half the total length, and if the width be twice 

 line «, the head is as wide as long. The labruni can not be included 

 in the measurement as it is not fixed. 



Foi" the removal of the genital segments the parts must be soft 

 and flexibU'. In cabinet specimens hot water swells tlie soft jjarts. 



