APBIL, 1906. THE .CHEMICAL NATURE OF SOME INSECT SECRETIONS. 29 



what amount of formic acid is present in different forms, and if 

 the amount is constant for each species. 



The ant selected for prehminary experiments was Formica 

 fusca, var. gnava Buck., as it was evident from its strongly acid 

 odor that formic acid was here present. 



A counted number of the ants were collected in weak ethyl 

 alcohol, and then subjected to distillation in a small flask with 

 steam. After the distillate had ceased to show an acid reaction, 

 a few cubic centimeters of dilute sulphuric acid were added to 

 the flask and the distillation continued, in order to ascertain if 

 any of the formic acid in the ants had combined with any small 

 particles of calcareous earth that might have been present adher- 

 ing to their bodies or in the alimentary tract. After removing any 

 trace of sulphuric acid from the second distillate b}^ means of 

 barium chloride, it was tested for formic acid by the method 

 given below, but not even an acid reaction could be obtained. 



The first distillate was then titrated with semi-normal potas- 

 sium hydroxide and its acidity noted. The solution of potassium 

 formate thus obtained was then heated for several hours at loo C. 

 on a water bath with an excess of mercuric chloride solution and 

 the amount of reduced calomel weighed. This method with 

 mercuric chloride was found, however, to be very unsatisfactory, 

 as it gave widely divergent results with distillates whose titration 

 determinations were very nearly equal. The poor results thus 

 obtained were probably due to the presence of organic matter 

 other than formic acid, and perhaps in some cases to an incom- 

 plete reaction of the alkaline formate. But as no other acid, e. g., 

 acetic, butyric, hydrochloric, etc., could be detected, titration with 

 semi-normal potassium hydroxide was selected as the most 

 reliable method. 



The following table, compiled from a large series of deter- 

 minations, shows the amount of formic acid in a number of species 

 of Formicid?e. As the agreement is in most cases so close, only 

 a few determinations are given for each species, besides several 

 widely divergent ones that seem worthy of notice. 



