the larjije wooden purlins allomng sagging, besides casting much 

 more shadow than the one-inch iron purlins so extensively used 

 at the present day. The amount of purlin required when using 

 larger glass is scarcely more than needed in an\' house, for the use 

 of many purlins is an important factor in preventing breakage, 

 and consequently an economy in greenhouse construction. The 

 iron house is therefore much superior to the wooden, and the truss 

 system of construction is excellent so far as the elimination of 

 shadow is concerned. The greater use of iron in the simple wooden 

 house has been a great improvement, and the tendency today is 

 to use more and more iron. But the importance of eliminating 

 shadow producing material, or in other ways improving the light 

 conditions, is not in our estimation as fully comprehended as it 

 should be by either grower or constructor. An increase in the 

 number of purlins is far less objectionable than an increase in 

 sash. 



From various experiments with the eflfects of shade on carbon 

 assimilation made in our greenhouse we have been able to realize 

 more fully how sensitive a plant is to light and shadov/. Even 

 the slightest shadow afifects the process of carbon assimilation, and 

 this in turn afifects the development of the plant. As already 

 stated, photosynthesis is in general proportionate to light inten- 

 sity, and the growth and development of the plant are directly 

 connected with it. 



INVESTIGATIONS RELATING TO LIGHT IN GREENHOUSES 

 Methods employed in measuring Light. 



In obtaining our light records we exposed tubes of a uniform 

 size and quality of glass filled with certain chemical solutions 

 sensitive to light. After being exposed they gradually change 

 color, the degree of change depending on the length of exposure as 

 well as on the intensity of light. The solution is then titrated. 

 The figures given in our tables therefore represent the results ob- 

 tained by titration by the use of a burette graduated to 1-lOOth 

 c.c, but which can be estimated at l-200th c.c. The readings are 

 ptirposely expressed in whole numbers in all cases, although in 

 most cases they represent decimals. 



In some of our experiments we made use of a clock to which was 

 attached a device for exposing the tube at any time required. 

 This device was elaborated by Mr. Neil Monahan and proved es- 

 pecially convenient for early morning exposures. (See figure.) 



In certain of our experiments the observations were made to 

 determine the difference between morning and afternoon light. 

 It sufficed for our purpose to call morning that part of the day pre- 

 ceding 12 m., which is known as mean time, and not the actual 

 meridian time; and that following 12 m., afternoon. 



