SEEDS AND SEEDLINGS. 



19 



to fall until it makes the required angle. The bend shoild be round 

 and smooth ; the Bun sen flame is apt to give buckled bends (Fig. 10). 



Fig. 9. 



Fig. 10. 



Fig. 11. 



(e) Bend some Tubing twice at Rig 1 lit Angles so as to form 

 Three Sides of a Rectangle. When laid down all three sides must 

 touch the bench. 



(/) Make two Nozzles. Hold a piece of tub- 

 ing by both ends ; soften the middle, and pull the 

 ends slightly apart. Cut the tube through and 

 round off the ends. 



(g) Complete the Wash-bottle. Bend suit- 

 able pieces of tubing to form angles equal to those 

 seen in the wash-bottle in Fig. 11. Push them 

 through the cork prepared in (6), and attach a 

 nozzle by means of an inch or so of rubber tube. 



4O. Experiments. In making experiments, 

 sketch the apparatus used. Make notes of the 

 materials experimented with (name of plant or part 

 of plant, number, condition, stage of growth, etc. ) ; 

 the duration of the experiment, date, time of day ; 

 the external conditions (temperature, light-intensity, 

 humidity, barometer-reading, etc. ) ; the precautions which seem neces- 

 sary, and the sources of error which may spoil the results. 



Always make "control" or "check" experiments, using the same 

 form of apparatus, set up at the same time, but with one or other of 

 the conditions different, e.g. in darkness instead of light ; with the 

 plants omitted ; with killed instead of living plants ; with plants in 

 different stages of growth. Also make ' ' repeat " experiments, using dif- 

 ferent plants under similar conditions or the same plants at different 

 times of year or day, etc. 



If your experiments do not succeed, try again ; if they give dis- 

 cordant results, try to account for these and to think out a method for 

 a repeat experiment under different conditions, with special precau- 

 tions, or for making a new experiment altogether. In drawing con- 

 clusions, try to distinguish between probability and actual proof. 



For various experiments with plants you will need (1) a rough 

 balance for large weights, and (2) a finer balance for smaller weights. 

 For the first, cooking-scales with weights from 7 Ibs. to oz., or a 

 cheap spring-balance, will suffice. If you have not a small balance, 



