70 VIEHMEYEE. 



of which are provided with a crown of short, hair-like bristles. Edwards ^ 

 discovered on the caterpillars of Lijcccna pseudargiolus Boisd., that the 

 slit-like pores gave forth a drop of liquid which was quickly licked up 

 by the ants and positively earned off and that the tubes upon the twelfth 

 segment neither voided a drop of liquid nor did the ants pay any atten- 

 tion to them. 



In spite of the exceedingly clear and detailed desci-iption and differen- 

 tiation of the organs as given by Edwards, later observers have repeatedly 

 put forth the assertion that only one of the two pairs is functional or 

 that the ants lick the evaginating tubes. Both these latter statements 

 are certainly false in so far as they apply to the larvEe before the second 

 moult. Moreover, the caterpillars of G. cnejus show clearly the fluid- 

 producing pore upon the third from the last segment, if De Niceville 

 does not err. I do not doubt but that when more careful research is 

 made the pore will be found upon all larvee in which De Niceville has 

 indicated only tubes. 



For C pandava Horsf. and Polyommatus boeticus Linn., at least, I 

 am able to demonstrate both organs. Apparently the only exception 

 seems to be Guretis thetis Dru., but according to De Niceville they appear 

 not to be visited by ants, and their tubes are to be differentiated from 

 those of the mjn^mecophilous lycsenid caterpillars as well by their form 

 as by their habits. It is therefore equally as possible that we have to 

 deal in this case with a substantially different organ. 



The more important of the two organs are the secreting pores, for 

 they furnish the attraction for the ants; but what the relationship of 

 the tubes is appears still uncertain. One might consider them as de- 

 fensive weapons or at least as the rudiments of such, as signal tubes, 

 or as scent organs for attracting the ants. A partial explanation of 

 their relationship is given by the earlier stages of the caterpillars of G. 

 cnejus. It should be noted that the development of both organs is not 

 symmetrical, but that they are drawn toward each other. Edwards 

 appears to have studied the earlier stages of these caterpillars in order 

 to determine the existence of the myrmecophilous organs. He says, 

 "The outward openings (the tubes) and the orifice in [segment] 11 are 

 visible in the youngest larval stages, but till near maturity the larva has 

 no use for the tubes and can not emit the seci-etion." 



In the caterpillars of 0. cnejus the relationship is different. At the 

 time of leaving the egg it possesses neither pore nor tubes. After the 

 first moult the pore appears; after the second, the tribes. This agrees 

 perfectly with the observations of Scudder ^ on the caterpillars of Everes 



'' On the Larva5 of Lyeaena Pseudargiolus and Attendant Ants. Can. Ent. ( 1878 ) , 

 10, 131-136. 



=^The Butterflies of the Eastern United States and Canada. (1882), 2, 914 

 and 959. 



