No. 528] SHORTER ARTICLES AND DISCUSSION 749 



however, is white, while that of signaticollis is bright yellow — 

 the latter apparently the positive diameter. In the cross, the 

 larva* and adults were equally of the andecimlineata type, al- 

 though it was shown in another experiment that the larval and 

 adult characters were inherited independently, each of four 

 classes of larva' living rise to three classes of adults, with Men- 

 delian numbers. It was found perfectly possible for undccim- 

 lineata to have yellow larva 1 , and for signaticollis to have white 

 ones, when the right crosses had been made. 



Are we to explain H. 700 A by saying that the result was a 

 genuine hybrid, from the make-up of which was somehow re- 

 jected the superficial characters of signaticollis f 



Another class of experiments, described as "experiments in 

 synthesis" was productive of extremely interesting results. At 

 Cuernavaca, Mexico, a suitable spot was planted with the proper 

 food-plants, and stocked with 210 specimens of signaticollis, 

 "from a standard location about a mile and a half distant" and 

 354 specimens of nndecimlineata. These freely bred together, 

 giving in the first generation 4.518 signaticollis, 11,744 mid- 

 type, and 5.091 u tidcciinlincafa. Gradually, however, the sig- 

 naticollis increased at the expense of the others, until at length 

 it was the only form present. This looks like natural selection, 

 and it will be noted that it occurs in a region normally inhabited 

 by signaticollis. The experiment seems ideally perfect, except 

 for the stated circumstance that there was a standard locality 

 for signaticollis about a mile and a half away. From what we 

 know of the migrations of these beetles, 1 it does not seem un- 

 likely that the colony was overrun with immigrating signaticollis. 



At Paraiso, in the foot hill rain forest, a locality was planted 

 with 100 each of signaticollis and undrcimlineata. This. I take 

 it, is in the undccimlincata country. The result was exactly the 

 reverse of the experiment just described — or not exactly that, 

 for the signaticollis practically disappeared even in the first 

 filial generation. Many other particulars are given, and there 

 is an important theoretieal discussion, but of course Professor 

 Tower has still in hand a great mass of unpublished material of 

 great value. One may hope that its appearance will not be too 

 long delayed. 



1 For example, in six years I had never seen L. decemlineata (the potato 

 l inn!.. r~. living. The tomato plants in my garden swarmed with them. 



