chiefly, necessary recharging but also noistvire contributed to long 

 nnd Bmioying, interruptions e 



To keep the raft GQ.oat 28 barrels vrere placed between the 

 tinl ers on both sides and foiu* additional ones uyider the floor of 

 the shop. At .■ least every other day these barrels would have to be 

 inspected and, if necessary, emptied by puiapihg. 



In general the tiEibers used for fraidng.vrere of generous di- 

 mensions,.. a provision vj-hich later proved to be usefiJl. The tir.ibers 

 rumiing lengthwise were 11 x 15-p: cm. as were the main braces -Thile 

 those upon vrtiich the shou stood measured 11 x 20 ck» The' v/alls of 

 the tallies were of 22 ram, boards. PujrtherMor e , n nuj-iber of cross 

 timbers iTere placed over the tanJvs for reinforcing purposes but 

 these Y;e;:e easily reir.ovable in the event the tanlcs were to be 

 hoisted- flrom the watera The driveshaft above each tank could be 

 discomiected by means of a clamp coniiection. The entire asseT±ily 

 was built at Tholen and, v.'-hen completed, tovred to the sit'e of 



op or at ions •• 



<,'■■■ 



In 1920 various chap.ges vrere made and additional tanlis instal- 

 led to give added opportunity for experimenting. Originr lly the 

 scheme was to construct eight much si^ia.ller tanks rlongside the two 

 existing ones. JlJ-nds. were hovvever insiif-ficiont and we finally had 

 to limit ourselves to the co;'Etruction of four new tanks. Inside 

 dimensions of -.these were 1.4 x 1.36 x 0.9 'm. ond they vrero attached 

 .to the side.cf t.he r.aft (see figure 3). The reasons for building 

 thes.e small tanks were economy, easier ha.ndling and the fact that 

 it took much less time to stock thorn vrith .the' .required .numbdi* of 

 Irrvae. The latter item is of great advantage in that a fevj- seed- 

 lobstors cm in o.ne day produce for one tank, enough larvae all of 

 the sane age. Thus cannibalism v; as reduced' and it was' also possible 

 to see at one time into all corners of a small tan]:^ an advantage 

 the large tanks did not have. Construction vras somev^-hat different 

 from thr.t of the large tanks. For one, the. screened openings in 

 the bottom v.'ere o:ndttod since they not only became clog;^ed very 

 easily but an eventual break in the screens might "have escaped 

 detection. On the other hand, larger v/indov/s were cut into the 

 sidea'/alls, nrmely winJovrs 0.6 m.. wide by 0.5 m. high in the walls 

 running lengthY^ise and 0.3 m. by 0.5 m» in the other v.^alls. The 

 -screens were mounted on slides inside of the tarilcs ".Thile on the 

 outside solid wooden shutters were similarly mounted whicii rx-de it 

 possible to remove the v/ire screens for cleaning after closing the 

 apertures by, means of. the irooden shvitters. This same arran^;ement 

 vfas later applied to the large tanks. The stirring device remained 

 the-;sane with the- exception that instead of supporting thp shaft 

 xv. a tripod the end was placed in an iron 'sqcket on the bottom. 

 Motive pojTer was also transi.iitted through conical gears and to 

 prevent rainwater from washing grease off the gears and into the 

 tanlcs the gears were covered i.-.dth tin housings. 



130 



