IN THE DIVIDING ROOT-TIP CELLS OF THE ONION. 27 



required to be attended once every 3 or 4 hours during the entire 24; 

 but it worked, and that was the essential thing. Thus the three com- 

 partments maintained temperatures at 10°, 20°, and 30° C, respectively, 

 each with a fluctuation throughout the growing period of less than 1° C. 

 above and 1° C. below the standard set. All other environmental 

 factors, including lighting, were apparently very uniform in the three 

 chambers. The machine was set in a cellar admitting light from the 

 north only. In this room the temperature during the period of 3 weeks 

 in which the thermostats were used did not vary more than 2° or 3° C. 

 This aided the maintenance of constant temperatures in the three 

 chambers. 



The onions were sprouted in earthen quart crocks and were supported 

 by floating wooden frames so that only the root base of each bulb ex- 

 tended into the water. When onions were first grown, February 1916, 

 for the preliminary work, they sprouted most readily, but in August of 

 the same year, when the constant-temperature apparatus had been 

 built and was in working order ready to receive the onion, the sea- 

 sonal conditions under which this bulb normally sprouts, or can be 

 induced to sprout, evidently were past. In all 5 varieties of onions were 

 tried out, but after 10 days none sprouted, but this time was well spent 

 in learning to maintain constant temperatures. By the time constant 

 temperatures had been attained in the three chambers, it was found 

 that, after scoring them deeply, the small white onions of quite uniform 

 character, commonly found in the fall vegetable markets, could be 

 induced to sprout roots. (See p. 35.) 



SAMPLING AND COUNTING. 



As was seen earUer (see p. 22), at a temperature of 18° C. (preliminary 

 study) the whole sequence of these 10 active stages of the mitotic cycle 

 for the onion root-tips, studied during the approach to the natural 

 growing season, occupied approximately 3 hours. This, together with 

 the fact that the highest point in their mitotic activity appeared at 

 11^40™ a. m., suggested that the most appropriate time for samphng, if 

 one wished to cover a whole active mitotic wave, would be from about 

 10 a. m. until 1 p. m. This succession was, therefore, decided upon and 

 19 observation-instants were chosen, each 10 minutes removed from its 

 predecessor, beginning and ending as above suggested. It is clear that 

 a completer and more refined analysis could be made if the observation- 

 instants were less remotely distant from each other; but it was desired 

 to cover as large a portion of a whole mitotic cycle as possible and to 

 make the cell-counts per individual sample as great as possible; hence 

 the necessity, in the interests of accuracy, to continue the observation- 

 instants in a series 10 minutes removed from each other. \Vhether 

 this is really economy working for accuracy can be determined only 

 when the relative influences of various factors (previously mentioned) 



