Genie Control of Development 349 



which may increase when the still unpublished additional facts be- 

 come known. We shall try to analyze the facts in close connection 

 with the previous discussion, thus making many points clearer than 

 they would be otherwise. In a general way, the work deals with 

 dosages of the recessive fourth chromosome mutant cubitus interruptus 

 (ci), which was mentioned in connection with the Dubinin effect 

 (see I 3 C c cc). The mutant ci removes a segment from the cubital 

 wing vein. The effect can be classed according to the length of the 

 removed piece, thus giving a quantitative appraisal of the effect of 

 the mutant locus or loci. A number of alleles are known of which two, 

 especially, enter the present discussion. One ci^ is a dominant; another, 

 called by Stern +^, is in the usual genetic language a very weak ci 

 "allele" with only a low percentage of penetrance, which is rather 

 subject to temperature influence. In the usual terminology it should 

 be called, say, ci^. Stern treats it as a low-grade + allele, which in 

 essence is the same but complicates somewhat the description of the 

 results. There are, in addition, two different grades of normal alleles — 

 called by Stem isoalleles — meaning that they produce normal veins, 

 because there is nothing possible above a normal vein, but the grades 

 are distinguishable by a different potency in the hemizygous condition, 

 + "/— , that is, over a deficiency. They are called +^ and +^. All these 

 alleles can be studied in simplex condition, one dose, over a deficiency, 

 or in haplo-IV; in duplex, or two doses; and in triplex, three doses, in 

 a triplo-IV fly; and all of them can be combined in the different dosage 

 experiments. In addition, a number of chromosomal rearrangements, 

 R, are available which have a position effect identical with the ci 

 effect, and they also can enter the different combinations. 



Of the alleles mentioned, ci behaves in the way we have already 

 described for the bobbed alleles; that is, the greatest effect upon 

 venation is produced by one dose (the exaggeration effect), and with 

 more doses the vein gap becomes smaller and smaller, though three 

 doses, the maximum available, do not yet suffice to produce a normal 

 vein. Their action may be described naively as producing a smaller 

 amount of vein-forming substance than the normal allele. For ci^, the 

 low allele, this would mean a smaller amount than normal, but not 

 very much less; and for the lower of the two isoalleles, an amount 

 almost sufficient to produce a normal vein in the simplex condition. 

 The dominant ci^, however, behaves like the dominant Bar effect; 

 with increasing dose the vein is more abnormal, so that the effect must 

 be described in terms of inhibition of a reaction. 



For the sake of analysis, we must distinguish three sets of facts, 



