78 [April. 



to them. I cannot say that the ants obtain anything from this Bi- 

 persia or from the Bipersia hereafter described ; but I have seen ants 

 {JE^ormica flava) evidently trying to extract something from Ooccids, 

 and this too directly from their bodies, and not from " honey-dewed " 

 plants. 



During the year 1890 a Naturalist friend called my attention to 

 a colony of ants which he said were constantly visiting some "American 

 blight" on his apple trees, evidently for the purpose of obtaining 

 from them " honey-dew " or other liquid, which he said he had seen 

 them extract on one or two occasions. On visiting the locality, the 

 so-called "American blight" proved to be a species of JPseudococcus 

 not yet determined, and the ants of the species mentioned above. 

 There were very many of the Ooccids, all adult females, just about to 

 construct their ovisacs ; these were visited by a various number of 

 ants, whose modus operandi w^as as follows : —three or more generally 

 surround a single Goccid ; one of ihem, usually the first comer, would 

 place its front legs on the back of the Goccid, with its head in the 

 same direction as the latter ; it would then commence to stroke its 

 back with its front legs and antennae ; when tired of this, a second 

 ant would repeat the process, and so on ; each awaiting its turn. 

 Although I watched them for some time, I did not see them extract 

 anything from the Ooccids. Judging from what I could perceive, the 

 Ooccids were rarely free from the ants ; they appeared completely 

 harassed, and had evidently tried to protect themselves by wedging 

 their bodies into the smallest crevices that would admit them. Through 

 the constant visits of the ants the bodies of the Ooccids were almost 

 denuded of their mealy covering. Whether the ants behave in a 

 similar way to the Bipersicc has yet to be ascertained, and there is a 

 wide field for investigation. 



Below I append a further contribution to the details of the 

 species. 



$ adult ; form variable, short rounded -oval, under-side flattish ; above gibbo?e, 

 or nearly hemispherical ; antennae variable, the more constant form is of seven joints, 

 of which the 7th is nearly as long as the 1st, 2nd and 3rd together ; 3rd, 4th, 5th 

 and 6th shortest, and in length nearly equal. By dividing the long 3rd joint, and 

 erasing the last joint of the fig. 6 of the antennfe given on pi. ii (I. c), a correct idea 

 will be obtained. 



Almost the same variation has occurred in the antennae of Bipersia 

 pulveraria (see same pi., figs. 7 and 7a), and a still greater variation 

 in the antennae of Prosopophora dendrobii, Doug. (Ent. Mo. Mag., 

 vol. iii, n. s.,p. 207, pi. 3, figs. 1& and 1^&), which has no less than four 

 variations ; some of these latter are undoubtedly due to parasitic 



