PHYLOGENETIC ORIGIN OF TERMITE CASTES. 121 



An examination of the five castes of R. flavipes (Figs. 1-5, 

 6-8, 13-15, and Table I) shows that a gradation of characters 

 may be traced throughout the series. In the three reproductive 

 castes (Figs. I, 2, 3), beginning with the winged adult of the first 

 form, or its nymph of the first form which is practically similar 

 in structure, an increasing loss of size is to be noted in the wings, 

 in the brain, compound eyes, and frontal gland, and in the sex 

 organs. The darkly pigmented body and the aerial habit in 

 swarming are found only in the adults of the first form, less 

 pigment and a wholly subterranean mode of life in the second 

 and third forms. In the case of the worker (Fig. 4), and to a 

 greater extent in the soldier (Fig. 5), there is a gain as well as a 

 loss of characters, -the gains being manifested in the larger head, 

 pigmented in the soldier, and the larger mandibles and mandib- 

 ular muscles; the losses, in the body pigment, the wings, brain, 

 compound eyes, and sex organs. The frontal gland is merely 

 vestigial in the worker, but is highly developed in the soldier. 

 These castes might be interpreted either as gradations in a series 

 of continuous or fluctuating variations, or as a series of regressive 

 mutations, i. e., mutations formed by the loss of characters, 

 comparable to the series of mutations found in Drosophila, 

 Morgan et al. (1915), Morgan and Bridges (1916). Should 

 the former prove to be the case, then transitional or intermediate 

 forms between the existing castes should be expected, but it 

 must be remembered that mutations also may be arranged to 

 fdrm a structural series, even though they may not have origi- 

 nated in this order. 



Up to the present time rather few intermediate forms have 

 been described. The reason for this may be either that they 

 have been overlooked thus far, or that they do not exist. The 

 intermediate forms known at present are as follows: 



In colonies of Termopsis angusticollis Walker, Heath (1903) 

 describes fertile soldiers with wing buds, which produced "normal 

 progeny." The question might well be asked what would be 

 "normal progeny" under these circumstances? 



Soldiers with vestiges of wing pads have been noted by the 

 two authors in several species of Calotermes: C. occidentis Walker, 

 and two new species from southern Florida; and by N. Banks in 



