364 PERFECT SOCIETIES OF INSECTS. 



which one sex may be known from the other. Thus secondary distinc 

 live characters, such as the beard in men, and the breasts in women, are 

 produced at a certain period oflife ; and these secondary characters, in 

 some instances, are changed for those of the other sex ; which does not 

 arise from any action at the first formation, but takes place when the great 

 command, "increase and multiply," ceases to operate. Thus women in 

 advanced life are sometimes distinguished by beards ; and, after they have 

 done laying, hen-birds occasionally assume the plumage of the cock : this 

 has been observed more than once by ornithologists, more particularly with 

 respect to the pheasant and the pea-hen. 1 For females to assume the 

 secondary characters of males, seems certainly a more violent change, than 

 for a worker bee, which may be regarded as a sterile female, in conse- 

 quence of a certain process, to assume the secondary characters of a fertile 

 female. 



With respect to the variations of instinct and character which result from 

 the different modes of training the young bees that we are now considering, 

 it would not, I think, be difficult to prove that causes at first sight equally 

 inadequate have produced effects fully as important on the habits, tempers, 

 and characters of men and other animals ; but as these will readily occur 

 to you, I shall not now enlarge upon them. 



Did we know the causes of the various deviations, as to form and the 

 like, observable in the three kingdoms of nature, and could apply them, we 

 should be able to produce these deviations at our pleasure. This is exactly 

 what the bees do. Their instinct teaches them that a certain kind of food, 

 supplied to a grub inhabiting a certain dwelling, in a certain position, will 

 produce certain effects upon it, rendering it different from what it would 

 have been under ordinary circumstances, and fitted to answer their peculiar 

 wants. 



I trust that these arguments and probabilities will in some degree reconcile 

 you to what at first sight seems so extraordinary and extravagant a doc- 

 trine. If not yet fully satisfied, I can only recommend your having re- 

 course to experiments yourself. Leaving you, therefore, to this best mode 

 of proof, I shall proceed to another part of my history : but first I must 

 mention an experiment of Reaumur's, which seems to come well in here. 

 To ascertain whether the expectation of a queen was sufficient to keep 

 alive the instinct and industry of the worker bees, he placed in a glazed 

 hive some royal cells containing both grubs and pupae, and then introduced 

 about 1000 or 1500 workers and some drones. These workers, which 

 had been deprived of their queen, at first destroyed some of the grubs in 

 these cells ; but they clustered around two that were covered in, as if to 

 impart warmth to the pupae they contained ; and on the following day they 

 began to work upon the portions of comb with which he had supplied 

 them, in order to fix and lengthen them. For two or three days the work 

 went on very leisurely, but afterwards their labours assumed their usual 

 character of indefatigable industry. 3 There is no difficulty, therefore, when 

 a hive loses its sovereign, to supply the bees with an object that will interest 

 them, and keep their works in progress. 



There are a few other facts with respect to the larvae and pupae of the 



1 Philos. Trans. 1792, viii. 167. Hunter On certain Parts of the Animal (Eco- 

 nomy, p. 65. Latham, Synops. ii. 672. t. 60. 

 Keaum. v. 271. 



