54 



THE AMEBIC AN NATURALIST [Vol. LV 



generally adopted in the laboratory. Several points have been 

 improved since, so that the procedure at present is as follows: 



1. Use bananas that are thoroughly ripe or over-ripe. 



2. Peel the bananas and weigh the pulp (100 grams of pulp 



provides for about four culture bottles). 

 3. Weigh agar-agar, 2 per cent, of amount of banana. 



4. Measure as many c.c. of water as there are grams of banana. 



5. Add agar-agar to water and heat until the agar has dissolved. 



(Complete solution is hastened by the addition of a small 

 amount of fresh water soon after the boiling point has 

 been reached.) 



6. While the agar is heating, press the banana through a potato 



masher or a coarse sieve, and place in readiness the 

 bottles (which should have been previously washed and 

 also preferably steam sterilized). Get ready yeast (Magic 

 Yeast ground up) and paper (absorbent paper, paper 

 toweling cut into 4-fold squares 3" X 2") and cotton 

 (stoppers may be reused, but should be dry sterilized by 

 enclosing over formalin. Cotton stoppers are better if 

 made rather tight and covered with very soft cheese-cloth). 



7. Stir banana into hot agar solution, ilix thoroughly. Mix- 



ture should not be heated any longer. 

 8. With ladle and funnel pour about 50 c.c. of the media into 

 each half-pint or pint milk bottle. (The media should 

 be at least -J in. thick to stick well.) 

 9. Sprinkle top lightly with dry yeast. 



10. Put in contact with media a 4-fold square of absorbent 



11. Stopper with cotton. 



12. Use same day. Best to use as soon as cool. Not good after 



two days. 



Flies can be mated in vials and then transferred to the cul- 

 ture bottles at the end of the day. A little food may be kept 

 going" by the alcohol method for use in vials, for covering over 

 mould patches in culture bottles, and for refeeding stock 

 cultures. 



The distortions in the ratios that arise from mortality char- 

 acteristic of given mutants and combinations (1) can not be 

 eliminated by direct methods. Fortunately, a large proportion 



