J. E. M. Mellor 71 



In order to collect the flies which emerged, tlio heap was enclosed by 

 means of muslin supported on wires (see Figs. 3 and 4) so arranged that 

 all sides sloped to an inverted box. Inside this box a muslin funnel was 

 constructed by nailing muslin to the sides and pulling it down to the 

 centre by catching it between the bottom of the box and the lid of a 

 round tobacco tin. Thus all the flies were guided to a small skylight at the 

 end of the muslin funnel, over which was set a balloon trap. 



In order to find out whether manure, which had almost ceased to 

 ferment still had attraction for adult flies or nourishment for their 

 larvae, a sample of horse manure was taken from a heap when it had 

 ceased to ferment. Part of this was placed in a large biscuit tin, provided 

 with a very fine gauze-covered hole in the lid to admit air, into which 

 6G half-grown larvae of M. domestica were placed. On September 28th 

 8 pupae were found, from which only one M. domestica emerged. From 

 another portion of the sample a water extract was made by soaking it in 

 a bucket of water for 36 hours, stirring at intervals, straining the liquid 

 through coarse muslin and evaporating it over a waterbath to the con- 

 sistency of rather liquid porridge. A watch glass containing some of this 

 extract was then placed in a cage with 36 M. dowestica (9 c^* and 25 ?) 

 which had been fasted for a week. They took practically no notice of it 

 and were all dead in about 10 days. 



Summary of observations. 



The results of the three experiments may be summarised thus : 



1. In a loosely formed horse manure heap the maximum temperature 

 is reached during the first 3 days from the time the heap is formed, pro- 

 vided that the manure is not more than 24 hours old when so treated; 

 and subsequently a fairly rapid and steady fall takes place throughout 

 the heap. This may be influenced, to some extent, by external tempera- 

 ture (Chart I). 



2. In a tightly packed heap the temperature registered by the 

 superficial portions resembles the temperatures recorded throughout a 

 loosely packed heap. On the other hand, the deeper portions tend to 

 remain at high temperature for a long time and then to fall suddenly 

 (Chart II). 



3. A comparatively lower mean temperature and a quicker fall was 

 registered in experiment C than in either experiments A or B. In the 

 former case they were due chiefly to a corresponding decrease in the 

 temperature of the air, but in the latter case chiefly to the fact that the 

 mean temperature of the m.anure in experiment B was kept relatively 



