THE HONEY ANT. 19 



great majorit}'), males, and perfect females. There 

 are, however, often several queens in an ants' nest; 

 while, as we all know, there is never more than one 

 queen mother in a hive. The queens of ants are pro- 

 vided with wings, but after a single flight they tear 

 them off, and do not again quit the nest. In ad- 

 dition to the ordinary workers there is in some species 

 Si second, or rather a third, form of female. In almost 

 any ants' nest we may see that the workers diflfer more 

 or less in size. The amount of difference, however, 

 depends upon the species. In Lasius niger, the small 

 brown garden ant, the workers are, for instance, much 

 more uniform than in the little yellow meadow ant^ 

 or in Atta harhara (PI. II. figs. 1 and 2), where some 

 of them are much more than twice as large as others. 

 But in certain ants there are differences still more re- 

 markable. Thus, in a Mexican species, 3Iyrmecocystus,^ 

 besides the common workers, which have the form of 

 ordinary neuter ants, there are certain others in which 

 the abdomen is swollen into an immense sub-diapha- 

 nous sphere. These individuals are very inactive, and 

 principally as living honey-jars. I have described in a 

 subsequent page a species of Ganvponotus (PI. IV. 

 fig. 1) from Australia, which presents us with the same 

 remarkable phenomenon. In the genus Pheidole (PI. 

 II. figs. 3 and 4), very common in southern Europe, 

 there are also two distinct forms without any interme- 

 diate gradations ; one witli heads of the usual propor- 



' Wesmael, Bull. Acad. Boy. Bruxelles, vol. v. p. 771. 

 c 2 



