Eeactions and Products in Interspecific Crosses 155 



a fraternity of hybrids, doubtful cases being measured and tested by breeding; 

 the two methods combined give certainty of operation. 



When the Pj apparently entirely dominant decemlineata is inbred the larvae 

 produced from such matings show pure reds, reddish yellow, and yellow in 

 proportion based on the scheme of 1 red : 3 red yellow : 1 yellow, although the 

 individual fraternities are highly variable. The adult progeny produced from 

 these groups is complex, and only that from the red line of larvae is of interest 

 here, the others not being presented. The red larvae produce adults that are in 

 pattern and coloration pure decemlineata, but show in form-index wide varia- 

 tion and a bimodal polygon with two modes, one on 2.600, the other on 2.950. 

 Matings from these modal classes develop from the 2.950 form a pure line that 

 is stable at once for the character of form-index and uniform for other decem- 

 lineata characters. The group that centers about the value 3.600 is variable, 

 some giving pure-breeding lines, others showing wide deviation or biinodal 

 conditions, but in all the numbers are not constant and the ratios are not in 

 harmony with present neo-Mendelian ratio systems. 



When a race is isolated that has the form-index 2.950 as its constant modal 

 value and is crossed with pure decemlineata from the parent strain, the product 

 of the cross in Pj is the production of an array all decemlineata in character 

 that shows a bimodal condition, with the exception that the form-index values 

 of the modal classes of each modal group are different ; namely, 2.600, 2.600- 

 2.800, 3.150. The appearance of these types is erratic and the fraternities show 

 diversity in the numbers presented, but from the modes 3.600 and 3.150 pure 

 races are derived. Two of the modal groups are apparently homozygous, but 

 the other contains individuals which when mated give pure-breeding lines with 

 the form index 2.600, while others are heterozygous. I have presented this 

 series of reactions in plate 16, and also data of some fraternities showing the 

 irregularity of the array with respect to the numbers present. 



The two lines that originate from this Pj array give two different form-types 

 of an organism that is decemlineata in all other respects, and I have not been 

 able to derive from either of these lines any characters that indicate the presence 

 of the oblongata parent. The race developed with the modal value of 3.150 is 

 perhaps the most interesting to us in that in all respects save form it is typical 

 decemlineata, even to minor habits and reactions, as well as in color and pattern 

 characters. The odor of its secretions are characteristically decemlineata and 

 not those of oblongata, which, even to my senses are recognizably different. 



It is probable that we are dealing in this race with one of two possible con- 

 ditions ; either that the line is the result of the transfer of all the commoner 

 decemlineata qualities onto an oblongata form-base, or that the form-determiner 

 of oblongata has been dissociated from its complex and replaced in the pure 

 decemlineata complex, the normal agent for form determination of that species. 

 The latter seems the more probable, and indication of its probability is found in 

 the change produced in the pronotal pattern of the race, which gives an appear- 

 ance much like that found in pure oblongata. 



Testing of this race by crossing with both of the parent species has given 

 suggestive confirmation regarding its constitution. When crossed with oblon- 

 gata, the product in Pi is intermediate between the two in all respects. Por this 



