The ideal arrount of stock to grow per bed is 



being worked out in nursery experiments. icn -ing final 

 conclusions, densities shown in tc.ble on page 18 ere 

 representative, in my 'best ju%iBent, of the amount to sow 

 per forty-eight square feet. 



ter sowing, the seed is covered with fine 

 river sand. :.; .t ean ths in wheel 



...', /enl$ -B possible with a shovel. 



It .on 1 eve lit with ,, straight- . board shout 



t\vo ioc\. lQj^]M9i|0jpl&uecl b. . .L.n, h the same as 



in levelling the surf; : ?i newly laid cement nalk. 



Formerly the sand wes firmly pa died with t- tamper, hut 

 this : or-weighted roller, : iit;/-\;wo inches wide, 



was tried out, which not 01 the cost of $* 



operation nian^ hun , tut ^Iso ; the sand 



raore uniformly. It is noe( or our 1 soil, that 



this roller ' Ifl ot in aiameter, to avoid a 



rippling effect on the surface of the he us. 

 4 Pepth of Coyer: 



>-*B^roue experiments aery show that the 



th of sc.nd co"ver is vit-1 to the ^ ation of the ser 



In the ti-'ole iollowing are eoe of the re suite obtained from 

 experiments.- pth of sejiti cover* it is believed t) 



uneven d n the is due more to uneven depth of 



covering- ti.tn to uneven sowing. Just what covering will 

 give best results e experiments. 



