SAIDY DATE OF EGYPT. 31 



Comparing first the Dakhla Oasis with, the Mecca garden in Cali- 

 fornia, both points having an excess of heat above the requirements of 

 the Saidy date, we may note that Dakhla has higher annual means 

 (both maximum and minimum) by 3.5° or 4° F. than Mecca, but the 

 Dakhla curves are broader ones, showing much higher maxima for 

 March, April, and May, nearly the same as Tempe and Mecca from 

 •June to September, but with Mecca about 3° F. hotter in August. The 

 Mecca minima are above those of Dakhla for both July and August. 



Now, from the deductions relative to the temperature of Baharia 

 made in a previous paragraph we must presume that the tempera- 

 tures at Mecca more closely approximate those of Baharia than those 

 of Dakhla. What Table 1 does not show is the occurrence of actual 

 frost temperatures in Dakhla, but, quoting from the same authority, 

 absolute minima of 0° C. were recorded in both January and Feb- 

 ruary between 1905 and 1911, though the writer has been assured that 

 nothing more than light frosts are ever experienced there. 



At Mecca, on the other hand, there is an average of 20 frosty nights 

 in the year, and the lowest annual minima of the record for 10 years 

 range from 13° to 26° F. Fortunately, these extreme minimum tem- 

 peratures endure for only a short period, usually not exceeding 30 

 minutes, as shown by the thermograph trace, and the frost never 

 penetrates the growing center, though at 13° to 17° F. most of the 

 outer leaves were killed. 



Comparing the Dakhla records with those of Heluan, the mean 

 maxima at Heluan are found to be from 6° to 11° F. lower 

 throughout the year, the greatest difference being for the month of 

 May. The mean minima for the year are nearly the same for Dakhla 

 and Heluan, but the curves cross about March 15 and September 15 — 

 that is, Heluan has warmer nights in autumn, winter, and spring, 

 but the nights are cooler than at Dakhla from the middle of March 

 to the middle of September. Heluan is practically frostless, though 

 Gizeh shows minima in January as low as —2.5° C, or 27.5° F. 

 The harvest in tlie Nile Valley is prolonged to about two weeks 

 later than in Dakhla. Frost would not interfere with the Nile Valley 

 harvest before the new year, but the relative humidity at Heluan, 

 always higher than at Dakhla, advances to 58 per cent in September 

 and October and 62 per cent in November, while at Gizeh the hu- 

 midity reaches 72 per cent in September and 75 per cent in October. 

 It is evident that this upper Gizeh district has little heat to spare for 

 the ripening of a choice packing date, and it is probable that it is 

 practical cultural experience that has confined the field of culture of 

 this date to the territory from Gizeh southward. 



Our next inquiry should be, why this date, whether derived from 

 the Libyan Oasis or from Gizeh, has failed to mature fruit at the 

 Tempe garden. The annual mean maximum for Tempe is 85.7° F., 

 only 1° below that of Mecca, where both the Lib3^an and Gizeh Saidy 

 strains have ripened perfectly, some years early in September. The 

 Tempe maximum curve is above the Heluan curve throughout the 

 year, except for falling slightly below in November and December. 

 It is far above it, practically with Dakhla, during June, July, August, 

 and September. 



It is to the Tempe minimum curve that we must look for an expla- 

 nation of the failure to mature fruit of this variety. This curve falls 



