IfflTHODS 



In the Field 



All collections reported on here were taken with nets of one 

 types, ioe»5 l=nieter (mouth diameter) nets with body (front and middle 

 sections) of 30XXX silk grit gauze (apertures 0o65 x 0o65 KEoJa rear ssctiss 

 and bag of 56XXX silk grit gaux© (apertures 0o31 x 0o31 ramo)o Th© 30XXX 

 mesh makes up about 97 percent of the straining svsrface of th* net, and 

 the 56XXX mesh about 3 percent o Details of th© eonstruetion of th© a^t 

 and the method of hauling have been described in a previous report (King 

 and Deraond 1953) o 



In this study two general types of hauls were employed g hor^ 

 izontal., with three net? towed simultaneously? and oblique j, using a 

 single net« Horizontal hauls of about 1 hour's duration,, with sampling 

 at three levels g were carried out on Smith cruises 4 and 6o Oblique 

 haals of about 30 minutes- duration to a depth of 200 meters were used 

 on Smith cruises 10, 12 j, and 17 and tfenning cruise 8s oblique hauls to 

 200, 300j and 500 meters were used on Manning cruise 9o 



The amount of water strained during each haul was estimated by 

 measuring the flow past' a current meter suspended in the mouth of the 

 net and computing the total volume^ assuming that the flow rate was uni- 

 form throughout the mouth of the neto The depth of the net during hauling 

 was estimated by measuring at 2-minute intervals the angle of the towing 

 wire and the length of wire out and making the appropriate calculation^, 

 assuming the towing wire to represent a straight line in the water, A 

 75 or 100-pound streamlined weight was attached to the end of the towing 

 wireo 



At the end of each haul the net was washed down thoroughlys 

 concentrating the catch in the plankton sock or bag„ 3Jie collection was 

 then transferred to a l=quart jar and sufficdent: formalin added to approx = 

 imate a 10=percent solutiono Th© formalin was neutralized with borax 

 and a completed label was placed in th© jaro 



That this method of hauling is capable of producing repeatable 

 results is demonstrated by a test of ths experimental or sampling error 

 among pairs of consecutive 200=meter oblique hauls mad® on cruise 9 of 

 the Manning e A comparison of 16 pairs of replicat© samples (following 

 Snedecor 1946, po 44) having a mean of 27o0 cCo/lOOO mo*p gave a meen 

 sample difference of 0o67 eso/lOOG mo^^ which by th© "t" test was shown 

 to be not significantly different (P > 0o4) from zeroo The standard de- 

 viation of the differences was 2o07, indicating that th© two memberi^ of 

 a pair of replicate hauls would be within 8o28 eCo/lOOO mo^ in 95 percent 

 of the instances 8 



