568 Balls. — Temper attire and Growth . 
being spent in taking consecutive readings, with practical certainty of 
a result. This does not, of course, include the time spent in the prepara- 
tion of the culture. 
Improvements. Though the apparatus has served its purpose up to 
the present, it suffers from a great many faults, chiefly, however, faults of 
convenience. By the help of an expert glass-blower, and by the use of 
steam-resisting cement, it should be possible to modify the chamber so that 
a number of cultures might be set up under aseptic conditions and examined 
successively, avoiding the loss of time which now follows transference from 
tube to chamber. 
The heating apparatus also requires to be improved, as it is at present 
satisfactory for only rising or falling temperatures. 
Experimental Results. 
Key to Construction of Curves. 
Typical sample experiments have been taken in each class. 
Ordinates represent the actual length of the hypha, and not the growth per minute. The latter 
can be seen by inspection of column 5 of the experimental data attached to each curve. Growth per 
minute is given in addition on Curve 1 to indicate error of measurement. 
When the control-point on the hypha has been changed during the experiment, the control-point 
has also been changed in plotting the curve, so as to give a continuous line representing the position 
of the tip of the hypha. 
The curves had to be plotted fairly steeply in order to show clearly the falling off of the growth- 
rate. The temperature curve had to be on the same sheet, so that a glance along the ordinate would 
show the temperature at the time of growth-cessation. Consequently, in Curves 1 and 5, the growth- 
curve crosses the temperature curve, making an apparent confusion, which can be rectified by 
consulting the side indices. 
The graphic ratio of the units of time, temperature, and length is the same in all the curve 
except No. 4, where the graphic temperature unit is doubled. 
Curve 2 is half the scale of the rest, but with the same graphic ratio. 
Curves 5 a and 5 b are plotted somewhat curiously in order to bring the graphic representation of 
growth-cessation in each to coincident points. This shows up the difference in growth-rate, time of 
appearance of the time-factor, and alteration in the stopping-point temperature more clearly than 
would have been the case with any other plotting. 
Small crosses on the temperature curves represent times when the galvanometer was at, or very 
near to, zero, i. e. when bath and culture were at the same temperature, and the error of temperature 
determination at a minimum. 
The Standard Cultures. 
The temperature relationships exhibited by freshly grown hyphaewhen 
supplied with fresh culture media-— the said hyphae not having been pre- 
viously raised above 20° C. — has been made the subject of some thirty 
preliminary experiments. The results are quite consistent. 
