26 BULLETIN 1105, U. S. DEPARTMENT OF AGRICULTURE. 



soil-moisture graphs show a marked depression in May and June, 

 a peak in July and August, and a secondary depression of varying 

 intensity in the fall months. Moisture is usually abundant from 

 December 1 to April 1. 



Successes and failures in reproduction during the past 11 years 

 may be readily accounted for by correlating rainfall and seed sup- 

 ply, as shown in Figure 2. The requisites for success are a good 

 seed crop, followed by abundant rain in July and August of the 

 next year, and an absence of severe droughts during at least one and 

 preferably two years after germination. 



A study of age classes in the vicinity of the Fort Valley Experi- 

 ment Station reveals a conspicuous absence of seedlings originating 

 between 1908 and 1914. An analysis of Figure 2 explains this. A 

 good seed crop was borne in the fall of 1908, and it germinated 

 abundantly during the rainy period of July and August, 1909. But 

 no rain fell during the 60-day interval from about September 10 to 

 November 15. A drought of equal duration occurred in May and 

 June of the following year. No exact counts are available, but it is 

 known that seedlings were abundant in August, 1909, and that few 

 remained a year later. Another good seed crop occurred in 1909. 

 Germination was fairly good in July and August, 1910. The Sep- 

 tember-October drought was not severe, but from April 1 to July 1, 

 1911, the rainfall was limited to a few light showers totaling less 

 than one-half inch. It is not surprising that few seedlings remain 

 from the 1908 and 1909 seed crops. 



The protracted summer and fall rains of 1911 gave rise to no seed- 

 lings, because no seed matured the preceding year. Here we have an 

 excellent illustration of the failure of reproduction with an abun- 

 dance of moisture, because seed was lacking. Germination and sur- 

 vival would have been excellent in 1911. With the absence of a 

 severe dry period in May and June, 1912, a large percentage of 

 seedlings would have survived. The rainfall of August and Sep- 

 tember, though too irregular for germination, was sufficient for seed- 

 lings a year or more of a^e. 



The light seed crop of 1911 germinated poorly, because of deficient 

 rainfall in July and August, 1912. On plots which had been arti- 

 ficially sown, a few seedlings came up in July, but died during the 

 drought of August and September. All that survived succumbed 

 during the 90-day drought of April- July, 1913. 



Had the seed production of 1912 been better, excellent reproduc- 

 tion would probably have resulted, because the abundant and well- 

 distributed rainfall during the summer and fall of 1913, and through- 

 out 1914 and the first half of 1915, favored a high survival. 



Another good seed crop matured in the fall of 1913. Abundant 

 moisture throughout 1914 resulted in good germination and survival. 



