32 Bulletin American Museum of Natural History. [Vol. XXIII, 



pseudogynes. Previous authors had been incUned to regard them as cases 

 of atavism, i. e., as reversions to the primitive female form (Adierz) or as 

 blastogenic anomahes, i. e., as due to some peculiarity in the structure of 

 the egg and hence referable to hereditary factors (Weismann, Forel). Was- 

 raann, however, took the view that they "are probably due to a post-embry- 

 onic sistence in the development of the typical female form, and arise from 

 Jarvse which were originally destined to become females, but had already 

 passed the stage of wing-formation and had then been converted into 

 workers." He conceives this deflection in the normal development to take 

 place in the following manner: ^ 



"The normal mating period of Lomechusa strumosa, according to my 

 ■observations, is from the middle of May to the middle of June. Usually 

 the Lomechusa larvae are to be found in the sanguinea nests only after this 

 period. At the very time when the Lomechusa; begin to lay their eggs, 

 ■sanguinea begins to lay a number of fertilized eggs which are to produce 

 workers. Both among these and the young larvse of the worker generation 

 the Lomechusa larvae, as I have observed, make terrific havoc. I have 

 noticed that their food often consists almost exclusively of these eggs and 

 young larvae, and that in the course of a few days a single Lomechusa larva 

 •can devour a great number of them. I have seen some of the larger Lome- 

 chusa larvse attack and devour several ant larvae 4-5 mm. in length, one 

 after another, although the Lomechusa larvae in this stage are usually fed 

 from the mouths of the ants. As there may be as many as several dozen 

 Lomechusa larvae in a single nest, the greater portion of the first annual 

 worker generation may be destroyed in the course of a few weeks, for the 

 parasites are extremely voracious and grow on the average almost a milli- 

 meter in length a day, and attain their definitive development (11 mm.) 

 from the egg to the adult larva (and that at rather low temperatures in my 

 observation nests) in 12 to 13 days. Hence there must arise a very per- 

 ceptible and sudden falling off in the development of the worker generation 

 and the ants endeavor to make good this deficiency by converting into workers 

 all the available larvae of the immediately preceding generation, that were 

 originally destined to become females. 



"Another factor, secondary in importance, however, compared with 

 those above mentioned, confirms the tendency of the ants to transmute 

 the female larvae into workers, namely the extraordinarily rapid growth of 

 the Lomechusa larvae. On this account the workers lavish the care, wdiich 

 would otherwise be devoted to the female larvae, on the Lomechusce instead, 

 for these impress the ants as being most admirable offspring. This in turn 



1 Die ergatogynen Formen, etc., loco cit. p. 632. 



