212 Relations betiveen Groictit and Encivoninent 



is .situated in a country district away from the fogs that are common to 

 urban areas the light conditions were good for any given period of the 

 year. This is an important point, as Gregory^, in experiments on cu- 

 cumbers at Cheshunt, near London, has shown that with high tempera- 

 tures light becomes a limiting factor, i.e. although other conditions are 

 favourable the growth is limited by the deficiency of light. Artificial 

 heating was installed and the minimum temperatures were kept above 

 freezing point throughout the winter months so that the results, so far 

 as they refer to the effect of minimum temperatures, are only apjilicable 

 in cases where the latter do not fall below 32° F. Throughout the period 

 the maximum and minimum temperatures were recorded daily except 

 on Sundays, but as the thermometers were read at 9 a.m., the Monday 

 readings included those for both the preceding days, thus insuring that 

 a high maximum or low minimum on Sunday was not lost. The weekly 

 numbers of hours of bright sunshine cover the seven days as they are 

 taken from the daily records made at Rothamsted. 



For the whole period of sixteen months the prevailing conditions of 

 temperature and light, as indicated by the weekly means of maximum 

 and minimum temperatures (Fig. 1), and the weekly hours of bright 

 sunshine (Fig. 2), are summarised in the following table: 



The mean maximum temperature is naturally greatly influenced by 

 the amount of sunshine, thus partly accounting for the large standard 

 deviation. The minimum is much less variable, as during the winter 

 months special care was taken to keoj) the niglit temperatures at a 

 reasonably high level by means of artificial heating. 



There is a close correlation between these three enviionmontal con- 

 ditions. The effect of each on the others depends upon the individual 

 characteristics of the house, such as aspect and ventilation and artificial 

 heating, and the extent of these effects can be measured by the corre- 

 lation, as follows : 



r^j^ ^ + -6284 ± -0384 



rfj^ = + -7560 ± -0271 



r^^, = -f -2895 ± -0595 



' 1917. Tliird Ainniul lUport of Exjicritiicitldl and Ji'tsain/i iSlulli/ii, C/u-^liiitil, JIiiIk, 

 ,,. 22. 



