Septembeb 2, 1904.] 



SCIENCE. 



311 



belongs is a primitive group, and breaks Lub- 

 bock's rule by retaining both stings and 

 cocoons. Its method of pupation is, there- 

 fore, of special interest, and has been recently 

 observed at Victoria, Texas, by Mr. G. P. GoU, 

 whose summarized report reads as follows : 



August 1. 5 P.M. Another larva is ready to 

 pupate and two keleps are industriously attending 

 it. One seems to. be cleaning the larva, while the 

 other is depositing earth around its head. 



7:30 P.M. Six ants are now covering two larvae 

 with earth, while the latter are continually 

 squirming about and disturbing the earth thus 

 piled over them. 



10 P.M. One larva is completely covered and 

 the other almost so. 



11:25 P.M. Two ants are removing the earth 

 from the first larva covered, showing that it has 

 spun a cocoon. The other larva is completely 

 covered, and third is being worked upon. 



11:35 P.M. The first cocoon has been taken 

 to another chamber and the particles of earth 

 are being removed. 



10:10 next morning. The third larva which 

 was being covered at 11:30 last night has finished 

 its cocoon, and has been carried away. This is 

 the fourth larva that has pupated since 4 P.M. 

 yesterday, in eighteen hours. 



From these and other observations the following 

 facts and inferences have been reached: 



The larva is entirely covered with earth when 

 leady to pupate. 



This earth is necessary as a basis for the cocoon. 



The squirming and apparent objection on the 

 part of the larva is an instinctive action to keep 

 the earth from being packed too close around it, 

 and thus not give it room enough to spin. 



all the members of the species; otherwise this 

 habit could scarcely have remained so long un- 

 known. The occurrence of the kelep in Mexico 

 is rendered somewhat doubtful, moreover, by the 

 fact that the Mexican Ectatomma ferrugineum 

 Norton, which Forel treats as a synonym of E. 

 tuherculatum, may prove to be a distinct species. 

 The National Museum has specimens from Mexico 

 which agree well with Norton's description and 

 figures, but offer appreciable differences from the 

 keleps. The habits reported by Norton for his 

 E. ferrugineum are also not those of the Guate- 

 malan ants. " This species is found only in the 

 encinales, or oak forests of the hot and temperate 

 region, where it lives in little societies under the 

 trunks of fallen trees." 



Time required to cover larva with earth, about 

 six hours. 



Time required for the larva to spin its cocoon 

 so as to permit removing to another chamber, one 

 and one half hours. 



The color of the cocoons changes with age from 

 a light gray to a pale reddish brown. 



Although the matter seems not to have been 

 considered by Lubbock and other investigators 

 of ants, it is easy to understand that, as Mr. 

 Goll says, the earth or some other material is 

 necessary to furnish support for the cocoon. 

 The naked larvse lying dbout on the level floors 

 of the chambers would have no means of sup- 

 porting their silk in the air, nothing against 

 which to spin. The majority of the lepidoptera 

 and other insects go into the ground to pupate, 

 with or without cocoons. Those which spin co- 

 coons above ground generally wrap themselves 

 up in leaves or seek crevices, corners or forks 

 of branches, across which their outer network 

 of threads can be fastened. It seems probable, 

 therefore, that this curious habit of building 

 earth cells for the pupating larvse is no re- 

 cently acquired instinct peculiar to the pres- 

 ent species or its immediate relatives. The 

 problem is as old as the social organization 

 and nest-building habit of the ants, and the 

 usual variety of solutions may be looked for 

 among the many cocoon-making species. 



The extensive labor involved in helping the 

 young ants to pupate makes it easy to under- 

 stand why so many members of the group have 

 discontinued the process. The time used in 

 making the earth cell is in some instances 

 much greater than that reported by Mr. Goll. 

 One of our larvse was surrounded by a row of 

 pellets of earth for over twelve hours, and, 

 though lying quite still, was continually at- 

 tended by three or four worker ants, waiting, 

 as it were, for the final emergency. The use 

 of earth in pupation constitutes a further rea- 

 son why the earth and stone nests built into 

 glass jar cages,* like those in which the ants 

 were brought from Guatemala, afforded a bet- 

 ter method of handling and observing them 

 than the horizontal glass plates or plaster-of- 

 Paris cells hitherto generally employed by 



* ' Habits of the Kelep or Guatemalan Cotton 

 Boll Weevil Ant ' Bull. 49, Bureau of Entomology, 

 U. S. Dept. Agriculture, p. 6, Washington, 1904. 



