1907.] Wheeler, The Polymorphism of Ants. 35 



should often occur is not surprising when we stop to consider that much of 

 the time and energy of the workers of an ant community are consumed 

 in other duties besides feeding the brood, such as foraging, excavating, 

 cleaning one another, etc. Then, too, the rate of reproduction is enormous 

 and must often out-run the available food supply, which itself is by no means 

 constant. The very slow development of ants in their larval stages is evi- 

 dence of slow metabolism, and as this cannot be due to low temperatures, 

 at least during the summer months, we must suppose that long periods of 

 enforced fasting or positive starvation not infrequently intervene in the lives 

 of larval ants. Still further conditions which may, perhaps, conduce to the 

 same result may be found in the apparent absence of a very definite and well 

 organized system of feeding the enormous brood, and the fact that this 

 important function is frequently entrusted to the presumably more or less 

 inexperienced callows. 



B. The North American Xenodusa. 



The Lomechusini are represented in North America by the single genus 

 Xenodusa. Our best known species is X. cava Leconte (PI. Ill, Fig. 41), 

 a deep red beetle, 5-6 mm. in length, with slender legs and antennae and 

 tufts of golden hairs (trichomes) on the abdomen, which is concave above, 

 like the thorax, and turned up at the tip. This species, which is so closely 

 allied to the European forms that it has been placed successively in the genera 

 Atemeles and Lomechusa, is stated by Wasmann to be "ziemlich haufig" 

 in the United States, but it has certainly proved to be decidedly rare in 

 my own experience and that of several coleopterists of my acquaintance. 

 Messrs. W. Beutenraiiller and C. Schaffer, who have given much attention 

 to our beetles, tell me that they have never taken it, and there is only a 

 single poorly preserved specimen in the large collection of the American 

 INIuseum. There may, of course, be localities in which the insect is as com- 

 mon as Lomechusa strumosa in continental Europe, but if such exist, they 

 have not yet been discovered or divulged. 



Leconte, who in 1865 first described X. cava under the name of Ate- 

 meles cavus, stated that he had never taken the insect and did not know its 

 host.^ Some years later, according to McCook,^ he succeeded in taking it in 

 the mounds of Formica exsectoides in the Alleghenies of Pennsylvania and in 

 the nests of F. rufa ( ?) in various localities. This latter datum, however, is 

 negligible, since the species of the e.v.secta and rufa groups in this country 

 had not at that time been clearly differentiated. 



' New Species of North American Coleoptera. I. Smiths. Miscell. Coll., No. 167, 1863, p. 

 30. 



2 Mound-makiiiK Ants of the Alleghenies, their Architecture and Habits. Trans. Am. Ent. 

 See, VI, 1877, pp. 253-296, pi. i-vi. 



