476 JOHN W. G O W E N 



2. Drosophila Genetics 



In determining the effects of irradiation on the germ cells of 

 higher organisms, as, for instance, Drosophila, the following points 

 are worthy of consideration. Eggs should be from a strain that lays 

 fertile eggs rapidly, at the rate of something like an egg every ten 

 minutes. The hatchability of the untreated eggs should be 95% 

 or more. Eggs should be laid in a limited period of time which is 

 standard for each test. Packard's standard method for such tests 

 is as follows: 



Fifty to one hundred females are allowed to lay their eggs on small pieces 

 of black filter paper wet with the juice of fermented bananas and spread 

 thinly with a little banana substance mLxed with yeast. In two hours they 

 will usually la_y enough for an experiment. At the end of the two hour 

 period the pajjers containing the eggs are removed and arranged in a conven- 

 ient way for irradiation. One hour after removal they are exposed to accu- 

 rately measured dosages of X rays and then placed in a moist chamber at room 

 temperature. These time intervals should not vary. Three hours after the 

 laying period commences, the eggs are in a most sensitive condition. As they 

 grow older, they become more and more resistant to irradiation, so that 

 three hours later they are approximately half as sensitive. This rapid change 

 in sensitivity makes brief exposures necessary. The eggs should be exposed 

 to X-ray beams having a relatively large number of roentgens per minute. 



About eighteen hours after exposure the eggs are examined under a binoc- 

 ular microscope with a 15 X magnification in order to count the number 

 hatched before the majority. These early hatched eggs were old when laid, 

 and therefore should not be included with those subsequently hatched, since 

 their sensitivity is very much less than that of the young eggs. The error in- 

 troduced by not subtracting these early hatched eggs is not large when the 

 dose is light, but it maj^ become quite important with a heavj^ dose. Two 

 days after exposure the sample is again counted and the proportion of the 

 hatching eggs determined. The corrected percentages are obtained by di- 

 viding the actual percentage observed by the per cent normal for tlie con- 

 trols, this l)eing the normal fertility of the strain under consideration. This 

 percentage should be 95% (jr better. If it is not, the material and fly strain 

 should be discarded and another one sought. This is not often necessary as 

 an inbred line of Drosophila varies little from year to year in its fertility and 

 hatchability. It is important to keep the fly strains used for these experi- 

 ments in stock bottles having a large excess of food and witliout overcrowd- 

 ing, as this ari'angement makes for good laying females. 



Experiments should be repeated on some good statistical design 



