BY VERA IRWIX SMITH. 509 



amount of moisture, so that tlie grass grows well, with thick rhizomes and closely 

 matted roots. 



On the Ttli November a small portion of this turf was dug up, and a search 

 made for larvae in soil spread out over a white concrete path. Together with 

 numerous Hymenopteron and Coleopteron larvae, and some cocoons of wasps, 

 and Syrphid flies, some twenty-eight larvae were found, of an undoubtedly Stra- 

 tiomyid type, ranging in size from 5 to 11 mm. (Text-flgs.l and 2) . Fourteen 

 of these were kept alive, and confined together in a small cylindrical glass pot, 

 containing an inch or two or soil, and some small grass sods, and covered witii a 

 glass lid. As the grass decayed, fresh sods were put in at intervals of two or 

 three weeks, and a few drops of water added with them, so that the soil was kept 

 just slightly damp. Five months later most of the larvae were still alive, but 

 showed very little increase in size. 



On the morning of tlie 13tli April a male Metoponia rubriceps was found to 

 have emerged. This date corresponded closely with that on which the tlies had 

 been observed first on the Museum lawn in the previous year. Accordingly, on 

 the following morning, a visit was paid to the spot from which the larvae had 

 been taken. Here, large numbers of imagines, both males and females, were 

 found already out, swarming above the grass in fairly rapid flight, the unusual 

 activity l)eing due, probably, to a period of warm sunshine following several 

 days' rain. From Mr. Hardy I learnt that a few individuals had first appeared 

 a week or two before, but tliat they had not become numerous until within a 

 few days of this time. 



Larvae were found to be jiresent in the soil in much greater numbers than 

 had been observed in the previous November. A rough searcii tlirough turf 

 taken from an area less than two feet sf|uare revealed eighty-five larvae, in all 

 stages of development, from lar\ae 3.2 nmi. long, to fully developed iiujiae in 

 larval skins 8 to 11 mm. long. Two female imagines were found just in the 

 act of emerging, and were killed and Hxed in this position, half way out of the 

 larval skin (Text-fig. 3). These were both taken just at the surface of the 

 soil. In the same position, among exposed rhizomes, were many empty caseS; 

 all showing a clean-cut, circular, aperture at one end, and measuring from 7 to 

 11 mm. long. In several instances the anterior extremity, forming a lid-like cap 

 to the case, was still lightly attached to it on the ventral side (Text-fig 4), but 

 broke away with the slightest movement (Text-fig. 4a). 



The vertical levels, in the soil, from which larvae of different sizes were ob- 

 tained, were carefully noted. The fully grown larvae were all found almost on 

 the surface, about the junctions of stems and roots, mostly wedged in between the 

 thicker rhizomes, especially of the paspalum grass. In colour, and seginented 

 appearance, these bear a rather striking resemblance to the larvae. From this 

 level, down to one to two inches below the surface, fifty of the larger larvae 

 were taken. At a slightly lower level, three to four inches down, among the 

 finer grass rootlets, were smaller larvae, always in close association with the 

 grass. Several were found attached l)y the head-capsule to roots, and one, removed 

 from the soil with its head buried in the root of a dandelion plant, remained in 

 this position for several hours. About eighteen inches below the surface, the 

 sub-soil and rubble foundation of the lawn were reached. No larvae were found 

 at, or below, this depth, and only a very few at the ten-inch level. These last 

 were of medium size, from 5 to 8 mm. long. 



