470 Scientific Proceedings, Royal Dublin Society 



The figures are given below : — 



Apparent volume of peas = 35 cubic iuclies. 



Volume of mercury poured in = 12 „ 



Actual volume of peas =23 „ 



Volume of interspaces 12 1 



Apparent volume of peas 35 3 



approximately. 



Volume of interspaces 12 1 • , i 



ir~. — i ^ j =?:?:= K approximately. 



Actual volume of peas 23 2 ^ 



In some of the experiments after the maximum pressure had been attained, 

 samples of the seeds were set aside to germinate. After a few days tlie 

 great majority of the seeds were sprouting, showing tliat tlieir experience had 

 not materially affected their vitality. 



A curious point observed was that a dark purple pigment developed in 

 the seed-coat of beans. The pigment is soluble in dilute acid, but is precipi- 

 tated as a brownish gelatinous substance in alkaline solution. Atkins (1) 

 refers to a purple pigment, probably identical, which develops in the seed- 

 coat of the bean when treated first with acid, then with alkali. The colouring 

 matter observed in our experiments developed in the coat of living seeds, as 

 deeply coloured specimens when set aside germinated freelJ^ 



The researches carried out by Coupin (2, 3, 4), Gain (7), and Atkins (1), 

 &c., some of which we liavo quoted previously, indicate that there is some 

 considerable variation in the amount of water absorbed by individual 



Coupin, and SchuU (16) in particular, lay stress on tlie fact that a large 

 proportion of seeds have injuries or defects in the seed-coat, which are 

 invisible, even on microscopic examination. 



For these reasons it appears to us that it would be advantageous to 

 estimate the pressure produced by a single seed. There are certain mechanical 

 difficulties in the way of such an estimation : in the fii'st place, the seeds must 

 be enclosed in a vessel, the sides of which are so rigid that its " give " under 

 pressure would be negligible; secondly, the seed must completely fill the 

 vessel, or else be packed in such a way as to leave it no free swelling space; 

 thirdly, water must be supplied freely to the entire surfaceof theseeds; fourthly, 

 fhe pressure exerted on the sides of the vessel must be measured, and yet as 

 little movement as possible be permitted. With these points in view we con- 

 structed the following apparatus shown in figure 2. It consists of two 

 sections E and G, made of cylinder cast-iron, which can be clamped by four 

 five-sixteenth bolts (two of which are shown at H). 



