766 EEPORT — 1887. 



little later I showed the fluid to Professor Wyndham R. Dunstan, who told me that 

 the characteristic smeU of formic acid could be plainly detected. This opinion was 

 further confirmed when it was found that silver nitrate was readily reduced by the 

 secretion.' In 1886 I obtained a larger number of larvse, and with the kind help 

 of Mr. J. P. Laws I was enabled to show that the secretion contains about thirty- 

 three per cent, of anhydrous acid. All the well-known qualitative tests were 

 applied to the secretion and to the alkaline salts obtained by neutralisinp: with a 

 standard alkali. Amonp: other tests the secretion was found to dissolve the oxide 

 of lead, a white crystalline salt being deposited. Although only a very minute 

 weio'ht of this was obtained, Professor Jleldola kindly offered to estimate the 

 amount of lead present in the salt. The weight was found to correspond to one of 

 the basic formates of this metal, formed by the action of the normal formate upon 

 the e.tcess of oxide. During the past summer I have had a very large number of 

 these larvffi, and the investigation has been continued with larger amounts of 

 secretion. The pipette has been applied for the removal of secretion between 500 

 and 600 times, and between 20 and 30 volumetric determinations have been 

 made. A mature larva which has not been previousl}- irritated will eject -050 grm. 

 of secretion, containing about 40 per cent, of anhydrous acid. Half-grown larvae 

 eject nearly as much, but the fluid is weaker, containing about 30 to 35 per cent, of 

 acid. The rate of secretion is comiiaratively slow ; e.ij.. two days and a half after 

 ejection two large larvte only yielded together -025 grm. of secretion. Two mature 

 captured larvae, to which the eggs of a parasitic Ichneumon had been affixed, only 

 ejected '035 and '045 grm. of secretion, having incompletely made up for the amount 

 lost during the attack of the hymenopterous insect. Starvation lessens the amount 

 of secretion and also decreases the proportion of acid ; but probably both these 

 efiects are due to jreneral health, and do not imply the direct formation of tlie acid 

 from the food. The different fnod-plants — poplar and willow — do not make any 

 difference in the amount or strength of the secretion. About half the total quantity 

 of secretion obtained was made use of in preparing a relatively large amount of the 

 normal formate, which is now in Professor Meldola's possession. The weights of 

 the constituent elements will be determined by combustion. [Since the above was 

 written. Professor Meldola has analysed the salt, and finds that it is formate of lead 

 in a state of almost complete purity, altliough he believes that a minute trace of some 

 other lead salt is also present.] The rest of the secretion has been used for other exact 

 methods of estimation and analysis under the kind direction of Mr. A. G. Vernon 

 Harcourt, the work having been conducted in his laboratory at Christchurch. Mr. 

 Harcourt suggested that it was most important to prove that the amount of acid shown 

 to be present by volumetric analysis is formic acid, and nothing else. This proof was 

 obtained in two ways. (1) A certain weight of the secretion was divided into two 

 parts ; the amount of acid in one of these was determined by the volumetric method, 

 while the other was decomposed by strong sulphuric acid, and the carbon monoxide 

 •which was evolved was exactly measured in the apparatus for gas analj-sis, and the 

 amount of formic acid present was calctilated from the data thus obtained. The 

 two percentages nearly corresponded, and as the latter was the higher it was ob- 

 vious that no other acid could be present. (2) A certain weight (•180 grm.) of 

 secretion was heated in a tube over a water-bath, and after drying at 100° 0. only 

 •0004 grm. of solid residue remained, and this was probably accidental. The rest 

 of the fluid was distilled into a tube containing carbonate of lead, and this was 

 afterwards heated to 100° C, and the water collected in drying tubes. As a result : 

 the increase in weight of the latter and the tube containing lead carbonate, the 

 weight of formate of lead obtained from the latter, and of stilphate of lead obtained 

 from the formate — all corresponded almost exactly to the weights which would 

 Lave been given by pure aqueous formic acid having the composition : water, 62^5 

 per cent. ; lormic acid, 37'5 per cent. It therefore appears certain that the secretion 

 consists of a strong aqueous solution of very nearly pure formic acid. 



' See Trans. Ent. Soe. London, 1886, pt. ii. June, pp. 1.5G, 157. 



