Eeactions and Peoducts in Interspecific Ceosses 149 



isolated from crossing with decemlineata, or if it did cross would retain the 

 purity of its type ; so that the conditions are not unfavorable for the origin of a 

 group in nature, the " origin of a species." 



I made the effort to put this hypothetical case to a test by the introduction 

 of L. decemlineata females into some of the habitats of L. diversa. I have not 

 thus far been successful, owing to practical difficulties in the control of the test 

 and the fact that L. decemlineata does not thrive in the habitats of L. diversa 

 in nature. In 1908 introductions were made at five places in the forested 

 valleys to the north of Cocomatepec, Vera Cruz, Mexico, where the food and 

 L. diversa were found in isolated locations. At these I saw the copulation of the 

 L. diversa males with the L. decemlineata females that were fresh from my 

 laboratory stock. I could not remain to observe the outcome, but when the 

 colonies were visited in August 1908 and again in 1909 no trace of the intro- 

 duced stocks was found. I saw the locations again in 1910, but found no indi- 

 cation of the success of the test. I hope at some future time to repeat on larger 

 scale and with better means this or similar lines of experimentation. 



The series of experiences with this cross and its products is highly suggestive. 

 The crossing of widely divergent types in the genus is of interest; one a tem- 

 perate savannah, the other a tropical montane rain-forest type with different 

 food-relations, the two differing in nearly all characters throughout; and only 

 in the general type of structure and in the large relations of parts and in general 

 pattern plans do the species compare, as is shown by the comparison of the 

 figures given of these types in Chapter II. The finding that less than 1 per 

 cent of the matings give progeny that are able to mature and continue the line, 

 and then only from one side of the cross, shows plainly that the problems pre- 

 sented in this combination are by no means solved. The lack of separation of 

 the F2 population into types of any sort shows that there is no gametic segrega- 

 tion or production of gametes of unlike kinds present. All indications show 

 that the gametes are like and permanent as far as experience goes in their com- 

 bination of the characters. The character of the strains that have arisen indi- 

 cates that the two parent types have each given to the new complex certain of 

 their determiner groups and that others have been lost or else are present in 

 inactive state or serving other purposes. 



All examinations and tests that have been made show that the type in body- 

 index is that of the decemlineata parent, and with this are the many character- 

 istics of structure and color-pattern, all those of the decemlineata type in pure 

 form. This, perhaps, indicates that the basis of the new combination has been 

 the gametic system of the female from which have been lost certain groups of 

 agents that have been replaced by other gametic agents from the male line. As 

 far as I have been able to discover, the limitation of food is complete, and 

 associated with this are a considerable series of reactions that involve the entire 

 body. The diverse catalitic agents that are engaged in the metabolic activities 

 and many sensory reactions are also distinctly present in the new type that came 

 from the male. The rate of development and many developmental details are 

 also those of the male line, so that the type is most complex. 



It is idle to speculate upon what has taken place in this combination, and only 

 further investigation can help in the solution thereof. It is clearly not like 

 any of the fixed heterozygous conditions thus far encountered. These have all 



