Drill Sampling Device for Fish Livers. Precision and Accuracy

Ind. Eng. Chem. Anal. Ed. , 1946, 18 (4), pp 269–271. DOI: 10.1021/i560152a016. Publication Date: April 1946. ACS Legacy Archive. Cite this:Ind. Eng...
0 downloads 0 Views 3MB Size
Drill Sampling Device Precision and

FA I ivers

A(

F. 8. SANFORD AND G. C oaltle Fishery Technological Laboratory, Fnh and

wvlliaiiro JOWICO,

x m e , Wash.

Because of the unhomogenoous nature of fish liven, error9 in vitamin A assays duo to sampling may b e largo unless an adequate number of corer are taken. In developmental work on a drill smpling device i n which over 2000 cans of soupfin shark and grayfish livers wore tested, 100 cares per ordinary commercial lot were sufficient to give vitamin A arrays that were reproducible within 5%. A n invortigation of tho accuracy of tho rampling showed that any bias, if present, war of small magnitude. Essentially tho same analytical results were obtained whether tho liven were fresh or frozen when sampled. The sampling war rapid. A hundred cores could b e taken in 15 minuter or less, and the rate at which tho livers were cored could b e varied within wide limih without appreciably affecting precision.

(Figure 1) consisting of a tube within which a tightly fitting, 10’ tating auger elevates the liver material into a sample bottle (1, 3). Figure 2 shows shark livers that had just been landed in the process of being sampled by a commercial processor with a sampler fabricated in his own shop. In the coume of the development of the sampler aver 2000 cans af soupfin shark (Galeorhinus zyrplerus) and grayfish (Squalw sucklii) livers were tested. The present paper reports the results of this work. A previous study (8) of a somewhat theoretical nature showed that taking a large number of corm is required for proper sampling and that, in general, i t is necessary to take at least 100 cores from a n ordinary commercial lot if the precision is t o be within +5%. However, conditions vary greatly. Not only am there differences of tex.. ... twe i n tl,e livers of various species such as aoupfin shark, grayfish, ling cod, halibut, .xnd tuna, h i i t tllc texture 4 5 0 varies n,iihin even I) single speclea. 1. Some Iww6 wall be f m h m d firm; , f t , nntl p r h n p s others will Le ~ i d c w . p ~ r i i : ~ ! tlmmposrri. y i n atlditiou, there may be other differenoes in the htributable to seasonal or geocauses, so that the number of IS involved in sampling is too allow far a rigorous determinrtprecision under all circumConsequently, in order to Figure 1 . Motor-Driven Core Sampler .e conclusions of the earlier cal studies, it has been necesSURPRISING fmture of the fish liver industry has bee u . sary to adopt an empirical procedure in the study of liver samthe praetioe of “blind buying”. Liversare bought from t bLe. pling and to limit thmc .experiments to livers landed in the fishermen without the benefit of a test for vitamin content. Th e Pacific Northwest. buyer may examine the livers visually (dark-colored livers usuall,P contain more vitamin A than light-colored livers), but since th e variation in the appearance of the livers is not commensurate with the variation in their poteucy, visual inspection will detect only differencesof large magnitude. This practice has persisted e v w though livers for some of the species, such as male soupfi0 shark, may vary in value per pound from 8.8 low &B 25 cents t o as high as $12 or $13, Obviously, liver trading on the hasis ctf “biind buying” is extremely hazardous. A primary difficulty interfering with sales on a potency basi has boon the problem of sampling. While hard-frozen liver can bc sampled with conventional core samplers, a method ha been iiecdcd whereby the “fresh” livers could be sampled jus 8 s they came from the fishing vessel. (If the livers are not froze1 salted, or otherwise preserved, they are called “fresh” by t h trade, even though they may actually be stale.) The fisherma are ordinarily a t sea for several days, ahd trips of 3 or 4 week: duration are not uncommon. When fishing conditions are favoi able, tho fishermen are anxious to put out t o sea again; undei sbandahly, they are not willing to delay long in settling thei r business. For this reason, the freezing of livers in prepara tion for sampling, while practiced extensi;*ely in sales betwee0 liver buyers and proeessors, has never been popular in primarY sales, since a delay of several days often occurs while the liver‘S are bcing frozen. To meet the need for a rapid method of sampling “fresh Figure 2. Motor-Driven Core Sampler in Use Some rhnk lircn NC exposed lo h o w n h r . 01 mlrrlal livers, this laboratory has developed 8 motor-driven core sample,r j

..._c_

-

A

?

269

.

*l.

j

270

INDUSTRIAL AND ENGINEERING CHEMISTRY

Vol. 18, No. 4

that it is beyond the scope of this laboratory to test them (One core taken per can) under all circumstances. HowLiver Deviation Deviation Oil Deviation Potency, of Liver ever, a fer$- experiments have of Oil ConPotency, of Oil PoPotency been performed with livers ofMillions Units from U.S.P. Units tency from Oil tent from Total Cans Sample Content, Average, T’itamin A Average, Vitamin A Average, from halibut, ling cod, rock Relative % Relative % per Lb. per Gram No. 7% Relative X in Lot cod, and sole, as well as ryith 16 1 72.5 0.96 13,600 0.74 4.48 0.00 halibut viscera. These lots 6.70 6.67 4.18 0.14 12,600 2 73.1 3 73.9 0.96 13.500 0.00 4.53 1.12 were all soft-frozen. While it 9.82 4.92 14,800 9.63 4 73.3 0.14 5 73.0 0.28 12,900 4.44 4.28 4.46 was not possible t o make an Av. 73.2 0.50 13.500 4.30 4.48 4.42 extensive test of precision, the 27 1 74.6 1.91 17,400 6.97 5.90 5.15 samples appeared t o be repre0.14 18,600 0.54 6.20 2 73.3 0.32 sentative, and no difficulties 0.14 19,100 2.14 6.34 1.93 3 73.1 4 71.6 2.19 19,400 3.75 . 6.30 1.29 were encountered. In Table 0.41 19,000 1.61 6.34 1.93 5 73.5 I11 are reportcd the results of .4Y. 73.2 0.96 18,700 3.00 6.22 2.12 sampling 35 cans of soft-frozen 0.39 2.33 1.70 7,710 66.6 1.04 40 1 tuna ( G e r m alalunga) livers. 0.84 7,620 1.55 2.35 68 0 1.04 2 7,620 1.55 2.34 1.27 These data show that the snm67.6 0.45 3 4 66.9 0.60 8,110 4.78 2.46 3.80 pler, a t least with frozen livers, 5 67.5 0.30 7,660 1.03 ’ 2.36 0.84 .I\,. 67.3 0.69 7,740 1.86 2.37 1.60 will work n-ell with species 1 other. than the grayfish or 56 69.4 0.73 5,220 2.97 1.65 1.78 2 69.6 1.02 5,420 0.74 1.71 1.78 soupfin shark. 3 67.9 1,45 5,640 4.83 1.73 2.98 A question has sometimes 4 68.7 0.29 5,220 2.97 1 62 3 57 Av. 68.9 0.87 6,380 2.88 1.68 2.53 been raised as to whether there 135 1 70.9 0.28 14,200 0.70 4.57 0.87 is any difference in the results 3.25 obtained when livers are sam2 70.6 0.70 13,900 2.80 4.46 0.00 4.66 1.08 3 71.8 0.98 14,300 pled frozen as compared to 4 70.9 0.28 14,800 3.50 4.76 3.23 Av. 71.1 0.56 14,300 1.75 4.61 2.11 when they a r e sampled “fresh”. Vsing soupfin shark livers, this was tested i n two instances: once with 17 cans, and once with 84 cans. The livers were first sampled I n forming a.sample for purposes of study, one core was taken before freezing and then resampled after they had been from each can in the lot, and the cores were then composited. frozen. As can be wen in Table I V in neither case did the difAdditional samples from the lot were formed in the same manner. Where the number of cans in the lot was small, the a s s a y tended to be some-xhat divergent. However, even in one cascl (Table I) where the lot contained only 16 cans, giving only 16 cores per Table IV. Results Obtained in Sampling Fresh and Frozen Soupfin sample, the greatest deviation from the average of 5 samples from Shark Livers this lot did not exceed 10%. I n general, as can be qeeii from Millions Total Tables I and 11, when lots containing a fairly large number of ran. o i l Potencv. U.S.P. L-nits TJnita X’ii cans were sampled, assays of both “fresh” grayfish and soupfin shark livers were reproducible to wthin 5%. 17 62.3 64.0 116,000 114,000 32 8 33.1 Livers and viscera are received in so many different conditions 84 61 2 61 8 97,000 97,600 27 0 27 4 Table I. Reproducibility of Samples of Fresh Grayfish (Dogfish) Livers

Table Table

II. Precision in Sampling 100 Cans of Fresh Soupfin Shark Livers

Sample No. 1 2

3

4

Av.

Oil Content, 7c

66.4 67.0 65.5 64.5 65.8

Table 111.

Sample No. 1

2 3 4 5

Av.

(One core taken Deviation of Oil Conte,nt Oil from Potency, Avera,ge, U.S.P. Units Relative T’itamin -4 per Gram 70 0.91 57,100 1.82 56,500 0.46 60,800 1.97 59,900 1.29 58,600

. per can) Deviation of Oil Liver Potency Potency, from Llillions, Averase, U.S.P. Units Relative Vitamin A 47, Der Lb. 2.56 17.3 3.58 17.2 3.75 18.1 2.22 17.6 3.03 17.6

Deviation of Liver Potency from Average, Relative 57 1.71 2.27 2.84 0.00 1.71

Precision of Samples of 35 Cans of Soft-Frozen Tuna Livers

Oil Content,

%

26.0 26.5 26.3 25.8 26.1 26.1

(One core taken Deviation of Oil Content Oil from Potency, Average, V.S.1’. ?nits Relative Vitamln A 70 per Gram 0.38 6,500 6,230 1 53 0 77 6,360 1 15 6,070 0 00 6,010 0.77 6,230

V. Accuracy of Samples of 100 Cans of Fresh Soupfin Shark Livers

per can) Deviation Liver Deviation of Oil Potenelof Liver Potency IIillions‘ Potency from U.S.P. from Averaoe Units Average, Relati% \-itamin A Relative 75 Der Lb. % 4.33 0.77 4.05 0.00 0 75 1.35 2.09 0.76 2 70 2.57 4 05 0 71 3.53 0.71 4 05 2,50 0.74 3.25

Samqle ho.

Oil Content,

1

66.4 67.0 65.5 64.5

2 3 4 Ground sample

%

(One core taken per can) DeviaDeviation tion of of Oil Liver Potency Potency of Core Liver of CGE Sample Potency, Sample from 1Iillioiis. from ~ ~ Ground U.S.P. Ground Sample, Units Sample Relative Titamin A Relati& 53 per Lh. R 4.8 17.3 1.1 5.8 17.2 1.7 1.3 18.1 3 4 0 2 17.6 0 6

Deviirtion Oil Content of Core Sample from Oil Ground Potency, Sample, U.S.P. Units Relgtive Vitamin .I 10 per Gram 3.4 57,100 4,4 56,600 2.0 60,800 0 5 59,900

64 2

Table VI.

~

60,000

~

~~

17.5

Accuracy of a Sample of 16 Cans Livers

of Fresh Grayfish

( 5 cores taken per can)

Oil Content, R Ground sample Core sample Relative yo deviation of core from ground sample

72.7 73.2 0.69

Oil Potency, U.S.P. Cnits Vitamin . I per Gram 14,200 13,300 4.92

Liver Potency, JIiilions U,g.P. Units 1 itamin .\ per Lb. 4.67 4.48 4.07

~

~

.

April, 1946

ANALYTICAL EDITION

271

the livers were buoyed up with oil. Nevertheless, the average results given by sampling the (g/a-inch and g/is-inch augers) Liver Deviation Oil Deviation Deviation thawed livers were the same Potency, of Liver Potency, of Oil of Oil as those obt,ained after they Afillions Potencv C o.. n.t.e.. n.t T7.S.P. Potencv DiamWt, SamOil from Bnits from ~.S:P. fromhad been refrozen. The analyof plingl ConAverage, Vitamin Bverage, Units Average, Sameter of Cores ses reported in Table V also ple Auger, per Sample, Time, tent, Relative A per Relative Vitamin Relative No. Inches Can Lb. $4in. 5% % Gram 46 A per Lb. % were made on a lot containing 1 0.79 10 63.7 0.31 74,500 3.76 21.6 3.85 appreciable free oil, yet the 71,400 0.55 20.5 1.44 43 63.1 1.25 5 3.66 2 a/( 3'8 1.53 20.7 0.48 0.78 70,700 A 2.95 10 64.4 3 )/'e results obtained by core sam64.2 0.47 70,500 1.81 20.5 1.44 J 7 . 8 6 . 36 4 B/is pling were not greatly different Av. 63.9 0.70 71,800 1.91 20.8 1.80 from those obtained with ground samples. Although the experience Table VIII. Precision of Samples of Grayfish Livers with frozen livers is limited, (One core per can taken by sampler with s/ie-inch auger) Liver Deviation sufficient work has been done neviRt,ion Deviation Potencv. of Liver . -.... to show that the sampler will Potency of o i l POM11liol;s of Oil Con- Oil Potency, ssmfrom tency from U.S.P. "nits Oil tent from U.S.P. Units piing Total be a welcome addition in this Average, Average, Vitamin A Content, Average, Vitamin A Cans in Sample Time, field also. Under present-day % Relative % per Gram Relative % per Lb. Relative 70 Illin. Lot No. 1.24 12,800 0.79 4.09 70.4 0.14 conditions, freezing facilities 8 1 2.23 2.36 4.13 13,000 70.0 0.43 8 78 2 are difficult to obtain, and 2.48 3.15 3.94 12 300 70.6 0.43 10 3 2.48 2.36 3.94 12:400 70.0 0.43 11 4 the livers are often held a t 1.24 0.79 4.09 12,800 70.4 0.14 13 5 temperatures too high for 1.89 4.04 1.93 70.3 0.31 12,700 Av. hard-freezing. Consequently, 15,100 2.58 0.81 4.88 71.1 1.57 1050 1 .. there has been a real need for 0.20 15,400 0.65 4.91 70.2 0.29 2 .. 15,700 1.29 4.92 0.00 69.1 1.29 3 .. a device which will sample 1 . 2 2 15,800 1.94 4.98 69.4 0.85 4 .. livers in the soft-frozen condi0.56 70.0 1 .00 15,500 1.62 4.92 Av . tion. The electric drills that 1.12 8,350 1.53 2.70 71.0 2.16 6 255 1 are equipped for holding drill 2.62 1.42 8,600 2.74 70.2 1.00 8 2 8,800 2.72 1.87 3.77 68.1 2.01 15 3 shanks up to 0.25 inch in 2.64 1.12 0.24 8,500 68.7 , 1.15 4 20 2.57 3.89 diameter generally do not 8,160 3.75 69.3 0.29 5 23 have sufficient power for use 69.5 1.32 8,480 2.17 2.67 2.10 AV. in sampljng soft-frozen livers. 2 . 0 1 9,140 2.54 1.20 61.3 0.49 256 1 20 1.23 2.50 8,850 0.40 62.3 1.14 2 34 However, those electric drills 0 . 7 8 2 . 4 8 1 . 2 0 8,890 61.3 0.49 3 45 that are equipped for taking 61.6 0.71 8,960 1.34 2.51 0.93 Av. 0.5-inch shanks are able to 0.84 1.75 1.3R AV. handle the greater torque a These livers were soft-frozen. Other values in table are for fresh livers. encountered with t,he frozen material. A skilled machinist can without difficulty make ference in the avera,ge results oht>ainedby sampling the livers in this change-over from the setup recommended for sampling fresh livers. the two states exceed 2Tc. In the study of liver sampling, the primary object, had been The standard sampler, employing a 1.123-inch auger, as described by McKee, Sanford, and Bucher (I), was used jn the to develop a method that would give reproducible assays of the vitamin .4 content. If it were later found that the absolute vitaabove tests. Smaller augers of 0,375-inch and 0.563-inch diammin content was not given hy the analytical method, and that the eter had been tried in earlier experiments ( d ) , and, for the information of those who might, consider use of such modified equipmethod was uniformly biased in one direction, a correction factor could be applied. In order to determine whether or not such a ment, results of the trials are presented in Tables VI1 and VIII. bias existed, t,he accuracy of the sampler n-as tested twice, once Precision was satisfactory, although clogging of the tube made with soupfin shark and once with grayfish livers. I n these tests, frequent cleaning of the auger necessary. samples were first taken xith the corer, and then the remains of The results summarized in Table VI11 demonstrate that wide the livers were ground. Three separate ground samples xere variations in the rate of sampling do not appreciably affect pretaken in each case to test their homogenei The three ground cision. These data also show the extreme rapidity with which the samples from each lot, gnvc~equivalent a ys well within the sampling can be carried out. Ordinarily, 100 cores can be taken of the method of analysis. Therefore, any deviation limits (2c0) within 15 minutes, so that more time is used for the incidental operations of removing and replacing the can lids than for the of the assay of the core sample from that of the ground sample can be regarded as evidence of a defect in the core-sampling method. actual sampling. The ease of use and the satisfactory results obThe results of these two tests, recorded in Tables V and VI, tained have led to the adoption of the device by many firms enindicate that the core samples may be slightly high as regards oil gaged in the fish liver trade. content and slightly low as regards oil potency. Further tests LITERATURE CITED would have to be made to establish this point. Fortunately, these two discrepancies tend to neutralize each other in the calcu(1) McKee, L. G . , Smford, F. B., and Bucher, G. C., FzsherU Market lation of the liver potencies, and if any bias exists, it is of small News,6, No. 11, 6 (1944). ( 2 ) Sanford, F. B., and Bucher, G. C., rnanuscript in preparation. magnitude. (3) Sanford, F. B., Bucher, G. C., and Stansby, M. E., Fishery AfarI n many cases, considerable free oil is to be found in the conkef X e m , 6, No. 1 1 , 9 (1944). tainers, especially if the livers have been long removed from the (4) Shockoy, C. F., and Sanford, F. B., Fishery Market News, 6 , No. 5, 9 (1944). fish or have been frozen and subsequently thawed. Free oil was encountered in the case of the 84 cans of soupfin shark livers P R E ~ E N T Eat D the Northwest Regional Meeting, AMERICANCHEMICAL recorded in Table IV, These had been frozen and thawed, and SOCIETY. OCtOBER 20, 1945. Table

VII. Precision of Samples of 55 Cans of Fresh Soupfin Shark Livers Taken with Sampler

~

a