590 



SCIENCE. 



[N. S. Vol. XIX. No. 484. 



clear, in the first place, that Castle is himself 

 not only lacking in ' critical caution,' but in 

 consistency, when he asks such questions. 

 Since he accepts the Dzierzon theory and pro- 

 ceeds to extend it to ants, he has no right to 

 change the theory during the transfer. All 

 adherents of this theory would agree in pro- 

 nouncing fertilization of worker bees by 

 di-ones an impossibility. They would contend 

 that this had never been seen. Hence if the 

 Dzierzon theory is to be extended to ants we 

 should consistently make the same assump- 

 tion. Nor would this be merely an assump- 

 tion. All observations — in this case far more 

 easily controlled than in the bees — go to show 

 that worker ants do not mate with males. 

 The case for Lasius niger is even stronger than 

 in the bees on other grounds also. In this ant 

 the difEerences in size and structure, both in 

 the reproductive organs and in the soma, are 

 vastly greater between the female and worker 

 than they are between the queen and worker 

 bee. No one, to my knowledge, has ever seen 

 even a receptaculum seminis in a worker 

 Lasius, though a very distinct vestige of this 

 structure is present in the worker bee.* But 

 even if the receptaculum were present, there 

 is no reason to suppose that it would function 

 any more than it does in the worker bee. 

 This would have to be admitted, however, if 

 we are to interpret Eeichenbach's and Mrs. 

 Comstoek's observations in accordance with 

 Castle's preconceived notions, for it is clear 

 that in the case of the Eeichenbach colony 

 months must have elapsed between the death 



* Castle's familiarity with the conditions in the 

 bee is well illustrated by his remark that ' dis- 

 sections of egg-laying workers, which were made 

 by Leuckart, revealed no seminal receptacle, hence 

 the eggs of such animals can not have been fer- 

 tilized.' The existence in worker bees both of a 

 vestigial receptaculum and of accessory glands was 

 pointed out by von Siebold more than sixty years 

 ago. Moreover, Leuckart, as he later admitted, 

 overlooked the receptacle in the dissections al- 

 luded to by Castle. A glance at the well-known 

 Leuckart and Nitsche chart of the honey-bee, 

 which can hardly be lacking in the Harvard 

 laboratory, would have shown Castle a by no 

 means insignificant receptacle in both the sterile 

 and the egg-laying worker. 



of the males each year and the laying of the 

 eggs, and Mrs. Comstock mentions the rearing- 

 of ' at least three complete broods' of workers 

 in the absence of males. From what we know 

 of other ants we could hardly suppose a 

 Lasius worker to function as Castle imagines 

 possible unless it were either a true or an 

 ergatoid queen. But no one has ever seen 

 an ergatoid female Lasius niger though this- 

 insect is not only the most abundant of ants 

 but the most abundant of animals over a large 

 portion of Europe and North America. In 

 this country it occurs in innumerable colonies- 

 from an altitude of 10,000 feet in the Rocky 

 Mountains to the sands of the Atlantic sea- 

 shore. I have myself collected and examined 

 thousands of these ants without ever seeing 

 anything that even approached an ergatoid 

 female. Is it probable then that two lots of 

 ants collected at random, like those of Eeich- 

 enbach and Mrs. Comstock, should both con- 

 tain fertilized ergatoid females indistinguish- 

 able externally from normal workers, especially 

 when we consider the remarkable propensity 

 of the workers of this and many other Formi- 

 eid^e for laying unfertilized eggs? Which, 

 then, is the more probable interpretation of 

 Eeichenbach's and Mrs. Comstoek's observa- 

 tions? Assuredly that which I advanced in 

 my former paper. 



Since the publication of my paper Professor 

 Forel has sent me a short article,* from which 

 I take the following paragraph: 



Zur Erkliirung des _ Polymorphismus der 

 Ameisen hat man zunachst die Analogic der 

 Bienen herbeigezogen, welche im Stande sind, in 

 den ersten Larvaltagen, durch veriinderte Erniihr- 

 ung vind Vergrosserung der Zelle eine Arbeiter- 

 larve in eine Weibchenlarve iimzuwandeln. 

 Ferner hat man nach Siebold stets angenommen, 

 dass die Mannchen aus unbefruchteten Eiern, die 

 Weibchen und Arbeiter dagegen aus befruchteten 

 Eiern stanimen. Letzere Thatsache schien auch 

 bel den Ameisen zu stimmen, indem ich selbst 

 and dann auch Andere stets Mannchen aus unbe- 

 fruchteten Arbeitereiern erzogen hatten. Doch 

 hahen die neuesten Untersuchungen Beiclien- 

 baoh's klipp und hlar den 'Nachweis geliefcrt, 



* Ueber Polymorphismus und Variation bei den 

 Ameisen,' Zool. Jahrh. Suppl. {Weismann's Fest- 

 schrift), VII., 1904. 



