124 FRANCIS E. LLOYD 



shelter ranged, on fog-days, about 10° higher at the maximum 

 than without. It is safe to infer that, were the prevaiUng day 

 temperatures 10° higher at the maximum, Phytolacca decandra 

 would grow and reproduce normally.^ and it may be remarked 

 too that the appearance and structure of the vegetative parts 

 would also be normal. It appears however that a full ten de- 

 grees is not necessary. 



In 1915 a plant growing outside but quite near to the glass 

 shelter produced on a branch quite close to the wall a single 

 spray of fruit. The inflorescence was so placed as to be sheltered 

 from the down-draft of cool air which is nearly constantly mov- 

 ing down the gulch in which the garden is situated, while it 

 received moreover reflected light and heat from the glass side of 

 the shelter. In 1916, another inflorescence was produced, but 

 was later blasted, after reaching full size and beginning to set 

 seed. It is naturally difficult to get at a precise expression of 

 the temperature differences to which this plant and those nearby, 

 and which did not produce fruits, were subjected. However, 

 by taking wet and dry bulb readings on typical days during the 

 warmest hours, the readings were obtained which are given in 

 table 2. 



Repeated readings at various times did not reveal any de- 

 partures from the above. It appears to be very near the truth 

 to say that, were the prevailing temperatures five degrees higher, 

 fruits and seed would be developed in at least sufficient abun- 

 dance to insure a population. 



This is further supported by the evidence obtained by de- 

 termining the differences of temperature shown by other ad- 

 jacent positions in which seed were and were not produced. Of 

 the 200 well-developed plants of 1916 above mentioned, one only 

 had red foliage. These plants, as is usual at Carmel, were 

 checked in their growth with the oncoming of the summer fog 

 after the latter part of May, and in consequence the more ter- 

 minal leaves formed rather close rosettes, generally with young 

 inflorescences more or less inclosed. In the red foliaged plant 

 only did these flowers open, and evidently started well toward the 



2 Phytolacca dioica fruits abundantly at Berkeley. 



