28o AGRICULTURAL EXPERIMENT STATION. 



small teeth on each side of the apex. The mandible (fig. 86, 5) 

 is normal and the maxillary palpus is short and broad, with a 

 moderate apical hair tuft and rather a large chunky, basal joint. 

 There are patches of hair on the surface of the palpus and some 

 of these hairs are feathered. 



The thorax is angular, wider than long, with moderate hair 

 tufts at each of the lateral angles and two smaller tufts on the 

 anterior margin. It is white in color, with a median black portion 

 becoming- wide anteriorly. 



The abdominal segments i to 6 are almost black, the anterior 

 two with lateral tufts oi four or five hairs each, the following 

 ones with two- hairs to each lateral tuft, diminishing in length 

 posteriorly. The seventh and eighth segments are white, with 

 short tufts only; the eighth with a regular row of five or six 

 scales on each side; the individual scales shaped as in figure 86, 

 9. The anal siphon is dark brown, almost black, about twice as 

 long as broad, slightly dilated near the base, the valves also' a 

 little dilated sO' as to flare at the tip. Tlie twO' rows of spines 

 consist oi five or six each, the single spines not curved, essen- 

 tially like figure 86, 10, though there is some variation. The 

 ninth segment is a little longer than broad, the same color as 

 the siphon. The ventral brush is moderate, with a few small 

 tufts below the barred area. The double dorsal tuft is short, 

 each part with one very long hair. The anal gills are long and 

 slender, tapering- to a point, and are without obvious tracheae. 



Habits of the Barly Stages. 



The suggestion that the species winters in the tgg stage is 

 made because it appears constantly associated with canadensis and 

 sylvestris, of which this is true. No very early collections have 

 been made in the Livingston Park woods, hence it is impossible 

 to say whether the adults taken June 23d were the first of the 

 year, or whether there had been an earlier brood. In collections 

 made July 2d young larvae of canadensis were found; but the 

 season of 1904 was not a favorable one for woodland pools and 

 there was no^ such abundance of wrigglers as occurred in 1903. 



The earliest actual date is July 29, 1903, when Mr. Van 

 Duersen collected full grown larvse and pupcE. From the pupfe 

 6 males and 4 females emerged July 30th, and 5 females 

 August I St. Some of the larvae pupated July 31st and adults 

 emerged August 3d, a period of three days. August 13th of 

 the same year another collection was made and half-grown lar^•ae 

 were found with full-gro\yn examples and pupae. In 1904 half- 



