RESEARCH METHODS IN STUDY OF FOREST ENVIRONMENT. 47 



fatal effects when the minimum light has been exceeded. For ex- 

 ample, if seedlings of a given species are grown with 20, 40, 60, 80, 

 and 100 per cent of the full available light, other conditions being 

 equal, and if the greatest accretion is put on by those having 60 and 

 80 per cent, while those having only 20 per cent barely exist, and 

 some of their number succumb, it will be fairly evident that the 

 optimum is between 60 and 80 per cent, and the minimum slightly 

 below 20 per cent for the given conditions of heat and moisture. 

 Both points may be found quite closely enough by curving the 

 growth data. Similarly, in other temperature and moisture series 

 different optima and minima of light may be found, and the abso- 

 lute optimum combination may be very nearly arrived at. 



On account of the difficulty of duplicating any set of conditions 

 at different periods, it is extremely desirable that the more im- 

 portant species whose relative tolerance it is desired to know should 

 all be treated during the same period, and also that an arrangement 

 should be effected which will make possible different combinations 

 of light with moisture and temperature. 



The following plan for such experimental determination of toler- 

 ance, while merely suggestive, may assist in initiating some work 

 along this very important line. The arrangements suggested should 

 accommodate about four species. It would, perhaps, be well to run 

 an initial test with rather gross differences in the light quantities, 

 as suggested above, and to repeat at a later date when the knowl- 

 edge obtained will permit more minute examination of the critical 

 points : 



Construct a solarium about 5-| by 8 feet, with its long axis lying 

 east and west, its floor and glass roof having possibly a gentle slope 

 to the south. The depth from glass to floor need not exceed 18 

 inches. Divide this into three equal parts by means of glass parti- 

 tions running north and south. If two layers of glass are used 

 throughout, having dead air between them, the purposes will be 

 more completely fulfilled without affecting light quantities appre- 

 ciably more than would the single layer of glass. Let the higher 

 north wall serve as entrance to the compartments, being closed by 

 a door whose inner surface has very poor reflecting powers. 



For each of these compartments 10 pans, each a foot square and 6 

 inches or a loot deep, will be required. These may be made of gal- 

 vanized iron with drainage openings in the bottom. Into each pan 

 put a measured quantity of soil, sufficient to fill it to within 2 inches 

 of the top. The pan and dry soil weight both having been deter- 

 mined, the amount of water necessary to maintain a given moisture 

 percentage in the soil may easily be computed, and this, added to 

 the gross dry weight, will give the weight which the pan should 

 -li'<\« a Her each watering. 



Each pan may now lie sown with sufficient seeds of the several 



peci< involved to inn. luce a good stand on the area of I square 



foot. Possibly I"" seeds of each species should he used in each, the 



