REPORT ON MOSQUITOES. 239 



times as long- as broad, with two rows of barbed spines, sixteen 

 to twenty-two in each row in the full grown larva. The single 

 spines (fig. 70, 9) have from one to three teeth. The ninth 

 segment is- longer than wide, with a barred area on the ventral 

 part, from which issue fourteen to sixteen tufts consisting of four 

 or five hairs each. On the dorsal apical margin are two tufts 

 of hairs, each with one very long hair. The anal gills vary in 

 length from very short to long, but are usually long as shown 

 in the figure. In the young larva they are always short. 



Habits of the Early Stages. 



Eggs are laid in places and as described for sollicitans. There 

 are more of them, however, and they are a trifle larger. On the 

 whole, under ordinai-y conditions, larvse are found nearer the 

 upland and rarely on the upland itself. In such cases they are 

 probably from exceptional females developing eggs after leaving 

 the marsh. 



As tO' the water, it may be fresh or salt or anything inter- 

 mediate, but, on the whole, fresh water pools formed by rains or 

 by drainage from the highland are preferred. In an exceptional 

 season like 1904, tide pools are just as well filled as the others. 



The eggs hatch just as readily as those of sollicitans, and 

 when young it is almost imposible to distinguish the two species. 

 As they become larger the maculate head of canator becomes 

 obvious, and then it resembles tccniorhynchus so- closely that 

 without a lens it is impossible to discriminate between them. 

 When full grown the shorter tube of the latter species and the 

 somewhat larger size of cantator give a basis for a distinction 

 which it recjuires experience to make. The pupa does not differ 

 obviously from the others with which it is found. 



As tO' the number of broods, that depends upon the weather; 

 but in midsummer, when the marshes become egg-covered, every 

 heavy rain or extra tide means a developing mass. The early 

 brood hatching in March develops slowly and irregularly, and 

 while adults begin to emerge early in April, it is not until the 

 beginning of May that all are out. Larvae of the second series 

 are not found much before the beginning of June, and they 

 mature during the last days of that month. Thereafter eveiT 

 pool will have larvse in some stage and every rain will make 

 some pools. A period of drought followed by a storm and ac- 

 companied b}^ a high tide will produce another general hatching, 

 and SO' on, as already set out in the p;eneral description of marsh 

 breeding. 



