148 Knapweed Gall-Jly 



the continental specimens of U . qnadrifasciata that I examined. Possibly 

 in this country U. solstitialis feeds also in the flower-heads of various 

 species of thistles, and its occurrence in these food-plants may have 

 been overlooked. 



U . quadrifasciafa differs from U. soJsfih'alis in the following charac- 

 ters. Its average size is smaller. The thorax appears almost black 

 owing to the very slight yellow powdering (Bestaubung) on the surface ; 

 all the femora are black also. The ovipositor is not much longer than 

 the abdomen. The four cross-bands on the wings are very broad, dark, 

 and sharply defined ; the first band is united with the second, and the 

 third with the fourth at the anterior border of the wing. 



Life-history, descripion of ovipositor, and metJiod of ovijiosition. 



A supply of galled flower-heads was gathered in September, 1912, 

 and one or more of these were examined from time to time in order 

 to find out when pupation commenced. The first indication of change 

 was observed on April 29th, the following spring ; in one gall, two 

 larvae were found whose skins were just hardening and changing to 

 a darker colour. On May 3rd, ten individuals were found in one gall, 

 and in nine of these the puparium was hard and definitely formed, 

 whilst the remaining one was still in the larval state. From this date 

 onward pupae were invariably found and in increasing numbers ; the 

 latest date on which I found unchanged larvae was May 29th. On 

 June 1st imagines appeared ; we may therefore say that the pupal 

 state lasts from four to five weeks. 



From June 1st imagines emerged every day until July 3rd ; after 

 this date very few emerged and only at intervals of several days, and 

 the last fly emerged towards the end of July. The period of emergence 

 thus extended seven weeks, but the greatest number of flies emerged 

 during the month of June. 



The flower-heads from which the flies were obtained were kept 

 inside a cool room during the winter, and in April they were taken 

 into the warmer laboratory. Outside in the open fields, development 

 would probably be slower and the flies would emerge later. In the 

 open, however, flies are on the wing in mid-June. 



The number of males and females that emerged between June 3rd 

 and July 3rd, 1913, was counted in order to ascertain the proportion 

 of the sexes to each other, lietween these dates 152 male flies and 

 130 females was obtained. In the early part of June, many more 



