By reference to table 65 it is seen that white and bar normally give only about 44 per cent of crossing-over in a two-locus experiment; lethal _sc_ then is expected to be situated at least as far to the right as bar. Females heterozygous for lethal _sc_ were therefore crossed to bar males, and their daughters were tested. The lethal-bearing daughters gave 144 cross-overs in a total of 1,734 males, that is, bar and lethal _sc_ gave 8.3 per cent of crossing-over. Lethal _sc_ therefore lies 8.3 units beyond bar or at about 66.5. The cross-over value sable lethal _sc_ was found to be 23.5 (387 cross-overs in a total of 1,641 males) which places the lethal at 43+23.5, or at 66.5. We know from other data that there is enough double crossing-over in the distance which gives an experimental value of 23.5 per cent, so that the true distance is a half unit longer or the locus at 67.0 is indicated by the 1,641 males of the sable lethal experiment. In a distance so short that the experimental value is only 8.3 per cent there is, as far as we have been able to determine, no double crossing-over at all, or at most an amount that is entirely negligible, so that a locus at 57+8.3 or 65.3 is indicated by the 1,734 males of the bar lethal experiment. To get the value indicated by the total data the cases may be weighted, that is, the value 65.3 may be multiplied by 1,734, and 67.0 may be multiplied by 1,641. The sum of these two numbers divided by the sum of 1,734 and 1,641 gives 66.2 as the locus indicated by all the data available. This method has been used in every case where more than one experiment furnishes data upon the location of a factor. In constructing the map given in diagram I rather complex balancings were necessary.
The fourth lethal which Miss Stark found (May 1914) in the inbred stocks of _Drosophila_ has not been located by means of linkage experiments. It is interesting in that the males which receive the lethal factor sometimes live long enough to hatch. These males are extremely feeble and never live more than two days. There is, as far as can be seen, no anatomical defect to which their extreme feebleness and early death can be attributed. { 79}
In studying the effect of hybridization upon the production of mutations in _Drosophila_, F. N. Duncan found a sex-linked mutation which he called "furrowed eye" (Duncan 1915). The furrowed flies are characterized by a foreshortening of the head, which causes the surface of the eye to be thrown into irregular folds with furrows between. The spines of the scutellum are stumpy, a character which is of importance in classification, since quite often flies occur which have no noticeable disturbance of the eyes.