014 



NATURE 



[July 15, 1920 



Note on the Habits of the Tachinid Fly, 



Sphexapata {M-iUo gramma) cornea. 



Fabre has given a graphic account of the patient 

 watch of this parasitic cuckoo-fly at the mouth of the 

 burrow of a species of Bembex, and of its cunning 

 in seizing- the moment wlien the " wasp " is half 

 within the .btirrow to deposit its tiny egg, pregnant 

 with disaster to the Bembex offspring, upon the body 

 of the insect victim intended lor the larder-nursery 

 wherein the mother Bembex 's hopes are laid. He 

 does xiot, however, appear to have witnessed in the 

 tragedy a phase that recently came under my notice, 

 and that is possibly restricted to, or perhaps only 

 easily observed, in cases where the foster-host carries 

 its prey along- the surface of the ground, or at best 

 flies only just clear of the ground. 



On the afternoon of June 22, when on one of the 

 heaths in this neighbourhood, I caught sight of a 

 black Fossor, Tachytes unicolor, carrying a paralysed 

 grasshopper. 1 followed, hoping to secure a photo- 

 graph of its operations at the burrow. Soon I 

 discovered that 1 was not the only follower, for at a 

 distance of about four inches there followed a small 

 Tachinid fly, which Mr. J. E. Collin has kindly identi- 

 fied as Sphexapata. conica. The fly followed the 

 *' wasp " with the utmost accuracy, maintaining its 

 distance with a precision that suggested a rigid con- 

 nection between the two insects; if the " wasp " flew, 

 the fly flew; if the " wasp " crawled, or indeed took 

 but a single step, the fly did exactly the same; and 

 always keeping distance accurately. For more than 

 fourteen yards — ^and there may have been many more 

 before I came upon the scene — did the fly thus follow 

 in the wake of the " wasp," until at length the 

 burrow was reached. The " wasp " at once entered, 

 leaving the grasshopper lying, belly upwards, at the 

 burrow's mouth; but before the owner was out of 

 sight the fly darted upon the grasshopper, without a 

 moment's delay deposited an &gg on its thorax, and 

 flew off — into my net. 



Fabre says nothing as to the distance at which the 

 fly stations itself when keeping watch at the mouth 

 of the burrow, nor of the interval between each in- 

 dividual when several "in a geometrical line " are 

 awaiting the critical moment; but the constancy with 

 which the fly kept station in the journey across the 

 heath, and the precision with which every movement 

 of the " wasp " was copied, suggested that at that 

 particular distance a clearer vistial image was secured 

 than at any other. Be this as it may, the fact is 

 worth consideration in discussing insect vision. 

 _ To this note I may appropriately add an observa- 

 tion made last year while watching an Ammophila 

 sabulosa filling in its completely stocked burrow. On 

 a stone close by there sat a small fly absolutely 

 motipnless, and apparentlv intentlv watching the pro- 

 ceedings. As soon as Ammophila had finished its 

 work and flown off, the fly leapt from its perch, 

 and at once began to scratch away the sand and small 

 stones in an endeavour to get at the larvae in the 

 subterranean larder. Fortunatelv. Ammophila had 

 packed its burrow too well, and the fly flew off de- 

 feated. This fly closely resembled Sphexapata conica, 

 but may have been an allied species. 

 _ Sharp ("Camb. Nat. Hist.," vol. vii., p. 509) men- 

 tions the fact that Miltogramma follows Hvmenoptera 

 carrying prey. Oswald H'. Latter. 



Charterhouse, Godalming, July 4. 



Temperature Variations at 10,000 ft. 



A SERIES of 500 aeroplane observations in North- 



East France in 1918-19 throws some light on the 



problem of temperature variations in the upper air. 



The correlation coefficient between pressure and tem- 



NO. 2646, VOL. 105] 



perature at 10,000 ft., taking all the observations 

 together, is 073. If the seasonal variations are 

 allowed for by taking the deviations from Mr. W. H. 

 Dines 's smoothed monthly means, the coefticient is 

 o-bg. The former value is higher, as the annual 

 variations of temperature and pressure in the upper 

 air are in the same phase. Both figures are rather 

 lower than the value 077 for 3 km. obtained by 

 Mr. Dines from balloon soundings, the observa- 

 tions being grouped in three-monthly pe riods. The 

 value 069 implies that a proportion v^i— 0-59'', or 

 72 per cent., of the standard deviation is still un- 

 accounted for. The partial correlation coefticient 

 between the temperature and the southerly component 

 of the wind velocity at 10,000 ft. (allowing for the 

 pressure) is 0-44, so that the southerly component 

 accounts for 10 per cent, of the temperature varia- 

 tions which' are independent of the pressure, or 7 per 

 cent, of the total variations. The effect of the west 

 component of the wind velocity is practically negligible 

 at all seasons. 



There are strong grounds for believing that a large 

 proportion of the temperature variations depends upon 

 whether the air supply was drawn originally from 

 the polar basin or the equatorial belt. This view is 

 supported by the humidity observations which were 

 made at the same time, as those for temperature. 

 For reasons set out in a paper which I hope 

 to publish, the original source of the air supply is 

 not very closely related to the wind velocity at the 

 place of observation, both polar and equatorial cur- 

 rents frequently following curved paths. This factor 

 of air supply operates in a very irregular manner, 

 with the result that the correlation coefficients vary 

 greatly from month to month. The coefficient con- 

 necting pressure and temperature at 10,000 ft. for 

 the period January-February, 19 19, based on fifty 

 observations, is as low as 0-09. In the winter 

 especially there are large fluctuations of the upper- 

 air temperature, the changes occasionally exceeding 

 30° F. within forty-eight hours both at 10,000 ft. and 

 14,000 ft. 



Mr. Dines gives a value o-86 to the pressure- 

 temperature correlation coefficient from 5 km. to 

 8 km., but this accounts for only half the temperature 

 variations. C. K. M. Douglas. 



Meteorological Office, Air Ministrv, W.C.2, 

 Jul y 8. 



The Brent Valley Bird Sanctuary. 



Sunday next, July 18, is the two hundredth anni- 

 versary of the birth of Gilbert White of Selborne, who 

 did more than any other of our countrymen to create 

 an interest in birds. The moment is therefore ripe 

 for an appeal upon their behalf, and for suggesting 

 how a fitting memorial to him may be established. 



The work which the Selborne Society has done in 

 the Brent Valley Bird Sanctuary, in the way of pre- 

 serving birds and testing nesting-boxes for use else- 

 where, is well known and has some considerable 

 value. The owners of the freehold wish now to 

 develop their estate, and if the money necessary to 

 buy the property is not forthcoming the sanctuary 

 will po. 



Matters have been made as easy as possible for 

 us, and we have been asked onlv 4500L for twenty- 

 two acres of building land which comes into the 

 London postal district. 



Mav L as chairman of the Bird Sanctuary Com- 

 mittee, invite the help particularly of those who are 

 fond of birds and of open spaces to save the wood? 

 Those who have been immediately interested in the 

 work have subscribed 300 guineas to start the fund. 

 Wilfred Mark Webb. 



The Hermitage, Hanwell, W.7, July 10. 



