THE EFFECTS OF HEIGHT, RADIAL POSITION, AND WETWOOD ON WHITE FIR WOOD PROPERTIES'

W. Wayne Wilcor University of California, Fnrrst Prdoct* Lahorntory, Richmond. California 94804

and

W. Y. Pong U. S. Frsest Service, I'acific Northwest Forest ;md Rongr IEspcrimmt Stittion, Pnrtlantl, Oregon 97308

ABSTRALT

Critical measurements and location of externally detrctahle defects werr. deternmiued for 20 whitc fir [Ahim c o n c o k ~ (C;onl. and Clmd.) Lindl.] trcrs. Hr,ards aawn from thew trees were identified with mgarcl to trce, height alx~ve ground, and radial position within the stem. Size :lnd locatinn of defc~.ts, and gmde and dlying sort of each hnnrrl were also rrrr,rtlcd. Moistnrc contont, spccific gr:~r,ity, tonghnt:ss, shrinkago, and liquid ahsorptio,~ were dvtcr- minrtl on s;~mplas taken at thror heights nnd thn:t: radial positions fnnn ripht of thc trrcs. T l ~ r data indicated that tlic sinker sort (thv onr requiring thr long~st drying timc) ramp primarily from thv ct.nte~. portion of the lowrat hvo l&ft logs, whilr t.hc mrky ( the one rrtp~iring the shortest drying time) and sap sorts were, msprrtivrly, from thr centcr and outer portions of l o p lthove the s~.cnnd I(j-ft log. Approxilnnt~ly 43% of the Ix,ard volume oanle from thr lowost 32 ft of stem. Ihr highest valoes nf moisture uontent, swcific gravity, t:ln~~nti:ll and radial tm~ghness, and tsn~ential and rotlisl shrinkage generally occurred in thc 1,nt.t log, dcrrc:~sing with I~cight, and in the outer thirtl of radial position, decreasing with npprnach tow,~nl the pith. I,<~ngitutlin:d shrinkage wns lowest at thr hutt and inrrp;isrd with Ilcight. Rec:~ose nrost typicnl w?hvorrrl symptoms occ~lrrrd in tllr centrr of butt logs, it i.i nrnrloderl that, eacppt for tltr nssocintiorl wit11 drying time, wrtwootl II:IS littlr or no detrimvntal rfftvt on thnso whit^ fir wootl propertips invrstigatt.tl.

INlXODU(T1ON

Wetwood (Hartlry, Davidson, and Cmn- dall 1961; Wilcox 1968) has I m n credited with possiblr responsibility for consiclrral~le loss in merclia~itablc volume of coniferous sawtimber (Alio 1966), but no datn are available to confirm this assumption. In an earlier paper on tlic properties of wetwood in white fir (Wilcox 1968) , it was reported that wood properties such as moisture con- trnt, specific gravity, toughness, and shrink- age varied wiclely depcmding upon the position in the tree from which the sample liar1 been taken, but therc appeared to be

little effrct of the presence of wetwood. Variation with height appc:ared to be more significant than that associated with radial position. By comparing the fundamental proprrtirs of tlie drying segregations into which wliitc fir boards are sorted during commercinl processing (Smith and Dittman 1960a; 19liOb) with tliosc dctcrmined on samples from known positions in the trce, it was inferred that drying segregation ("sort") also was highly clepndent upon location within the tree. The relationship could only be inferred, howevc~, as tlic identity of 11oards with respect to position in the tree normallv is not Anown. I t also was annarrnt . .

'Appreciation is extended to American Forest from tll&se data (wilcox 19fi8) that, wllile I'rorlucts Cnrp., Martell, California, for assistancr in board segregation was intcnded as an rsti- ol,taining the sanlplr material and for allowing thc stodv to h oerf~lnned in thrir to G. p. ,.jrnese. mate of drying time, it also provided an R. L:. ~ a y , j. I;. Wrrbe, D. C. D A I ~ ~ , L. A. ran; estimate of oilier wood proprties. and Mrs. Nancy D. Oldhnrn for t<vlinic;~l assistanrc Many problems ill the processing of white rltlring varilrlls phascs of the rrscarch; and to Ilrs. fir lumber appear to from dryillg W. h.lrKillap and C. L. Chiang ant1 Mrs. Mitrgarct R. Wilcor for ndvicc on thc statistical tmnlysrs. of boards Iiaving material that normally St~ppnrted in part I,y McInti~.r-St<~nnis. woulil he placed in two diffcrent drying WOOD ANT) wnIin 47 SI'HINC 1971, V. 31 1 )

48 W. WAYNE \VII.COX AND W. Y. I'ONG

sorts prc.sent in the same board; however, there are no data availal~lc to snbstantiatc this or to snggcst what portions of the tree sncli boarcls come from. Rccausr funda- mental propcrtics of the wood affect its rluality as a raw material and througl~out processing, and l>ecause bot11 fundamcntal pr~pc'rtios and proct*ssing segrc~gatioiis ap- prarrd to b r related to position in thtr trcc, a study was dcviscd to mcasure each of thcsc factors in a manner that would allo~v thc examination of tlic relationships bc- tween all factors. This paper reports t11c resnlts of a portion of that stndy.

MATERIALS AND METHODS

Twc~nty white fir [Abias conmlor (Cord. and Clcnd.) Lindl.] trccs wcrc s(~lcctrt1 for thc study from a single mixed conifrr site in El Dorado County, California. Tllc sclcc- tion arca, locatcd on the west sidr of thc Sicrra Nevada mountains at an elevation of 62(M ft, was a relatively young stand on a gond site in a prcdominantly white fir- incense cetlar-sugar pine timber type. Sam- plo trers wcrc representative of tlie range of diam~.tn.s, heights, and externally indicated drfrcts commonly found in trccs loggcd in this arca. Eight of thrsc, trecs wcrc srlcctctl as a rcprcscntativc su1,sct for which various fundamcntal properties wore dctrrminod at nominal liciglits abovr ground of 1, 34, and 72 ft. Thr diameters at brcast hcigllt (4.5 ft itbovc! ground) of the full set of sample trccs ranged from 17.2 to 46.9 inches, with a mean of 30.7; ages ranged from 79 to 176 years, with a mean of 131; and heights ranged from 88 to 182 ft, with a mean of 131.

In the ficsld, each tree was assigned a numbcr. This identitv was maintained

knots and defects were diagrammed (Jack- son, Henley, and Jackson 1963; Pong and Jackson 1971).

Prior to sawing, thc large m d of each log was paintrd in a target pattern with tlinx~ colors. The widtll of oach band of color covcrccl approximatcly one-third of tlie large-end radius of the log. T l ~ c paint color on tllc cnds of tlie rc>snlting boards drlin- cated, to the ncarcst one-third of the radius, thc radial position within the log from which the boards had n)me. All ~ilaterial cut from a given log was color-codcxl at thc lrc~adsaw as tlie log was initially hrokm down. From this code, crewmembers nn- mt>rically identified cach hoard with the appropriate numbcr of the log from which thr board was d(~rivec1. On thr grcan chain, each board was stamprd with a consrcutivc~ numbrr, visually sortecl, and then ~nilrked for drying sort 11y mill pcrsonncl according to prrviously ~ublislicd guidrlincs (Smitl~ w d Dittmim 1960a; 19601)), and then graded by a certified grader. Normally, mill practicc does not call for grading white fir in tlie green conditio~~. hut in order to provide a comparison of I~oartl quality 1,rfore and after drying (I'ong 1971), l~oards were graded oil the green chain as wcll. During dry-sorting, a photographic rccorcl of each hoard was automatically mad? on color film using a cin(~pu1sr mmera pulscd by an automatic pliotoolcctric triggering mechanism (Pong, Bass, and Claxton 1970). The film record contained for cacli I)oard the m d codor codc, scqoenti:~l 1)oard num- Iwr, log nutnbcr, drying sort, green and dry grades, pencil trim, dcfccts and othcr sur- face conditions, width, length, and thick- ness.

To dc~termine locations in the stem that tllroughont proccssillg.' A11 visible dcfccts give rise to sort, each was

visually rated to thc nearest 10% as to the and other trrc s~~rfncc charnctcristics \vcrc percentage of its end arca occupied by cacli

recordrd for tlic standing tree and again of the radial position color codcs. To insure after ft!lling and bucking. Tags, containing accllracy, this rating was jntlcl)cn(1cnt]y tllr identifying trrc and log nomhers, were ~ n a d r on all boards by two proplc. The stapled on tlic cnds of each log so that tlie percentage rating was then multiplied IJy log's position in thr standing trcc would thc nominal 1)oard-end arca to 11rrivc at a remain known throughout processing. At mcamrc of the board-end arca, in squarc tlie mill yard, racli log was carefully exam- inches, coming from each radial position ined on all sides and the locations of all (each H radius). Calcolations werc maclc

TABLE 1. h4con board-end arso inc1~e.v) ~p

H p i ~ h l itcrass ,111 sor1s illill nwinz mn sinkpr C D T ~ SOP mc~i i l ~ positillns --

"""','i~ Rndinl poritirln Pith Middle Outer Pith hlidrlle Outer Pith Middle Out?? hlpan rzr,

1Icigbt (113-i, 1 " ~ s ) Pllrnmetrr -. -

1 Mr:m 14.9 11.2 9.2 11.2 4.8 0 19.2 8.6 9.7 11,7 392 S;tmplc size 105 172 78 1 1 1 4 14 10

2 Alean 12.7 10.8 10.0 23.2 2.4 0 13.9 7.9 9.0 414 S~unplr sizr 68 132 $13 3 1 1 24 52 40

3 Mmn 12.7 10.3 9.8 15.6 10.(i 8.0 12.6 8.6 9.4 355 Samplc siae 23 53 44 12 14 1 49 84 75

4 M,:iun 2 I 9.9 13.9 11.2 4.0 10.2 10.8 10.4 2$13 Salnplc she 12 20 If3 32 38 3 28 71 71

5 Mean 0 0 0 13.7 9.8 3.3 9.3 9.8 10.0 S;~~nplc size 1 1 1 27 32 13 14 49 64 9.8 202

6 Mean 8.9 9 14.6 13.7 12.9 2.0 7.8 10.0 8.9 223 Samplc sizr 5 7 8 26 36 12 22 ~l 56

7 Mran 0 0 0 12.7 9.4 4.6 7.7 7.6 10.2 9,0 95 S;nnplc: size 1 1 1 15 21 7 7 17 25

Sinker C n r h Sap

Sort across all hriahts hlenn 11.2 11.2 9.8 and rnrlial positions Sw~ple size 842 297 835 - ~ ~ - -

Pith Middir O,,t<.l -. -- .

Ilarli.tl pwitioo across Mt?an 12.9 10.3 9.3 all hcights :and sorts S;~mplc size 479 867 628 Total snmplc sizc = 1974

in square inchrs of c?nd area rather than in board fcct, thcrchy ignoring hoartl lmgth for two reasons: ( a ) considcration of length as a variable would include taprr ( thr dif- ference between large-end and small-cnd log diameters) as a confounding variable, and ( b ) board length was hrld relatively constant throughout the experimental nm at a nominal 16 ft. Actual mean board lcngth for thc entire set of hoards included in the samplc from which these data wrrc rlcrivrd was 15.2 ft, which when roundcd to tlic nearcast 2-ft-lmgth class justifies an assumption of n constant 16-ft avcragc hoard Icngth. Tlicsc data for board end areas froiii each ritdial position were further classified by drying sort and height.

For stndy of funrlamental properties, a

disk approximatcly 15 inclirs long was rr- moved from ench of tlic eight trros at c!ach of tlic tlirrr samplc heights at the timc of bucking. At thc same tune another disk about 3 inches thick was also rcmovcd. From this disk 1)locks approximately 1 by 1 by 3 inches in size werp immediately split, sealrd in watertight, tared hottlrs nnd later used for dctcrmination of frrsh moisturc content. Tho mcasurcment for each 32ft log, the usual hauling length, was thcrcfont dis- placed up the stem npproximatcly 1'' ft by each srt of snmplc disks removetl.

Thc samplc disks wcrc prncrsscd and tcstrd in tlic samc manner as reported earlier (Wilcon 1988). Absorption was cal- culatcd from the same experimental data and in the same manner as retention in an

50 W. WAYNE WILCOX AND W. Y. 1'ONG

TARLE 2. Annlysi.9 of oariancc table for data rezwcrantrd in Table 1 - -

SDII~-?- SS rlf \8 P Q?

Sort 912.48 2 456.24 11.93**" 0.01 Flright 629.27 6 104.88 8.74* 0.00 Rndi;d position 3008.75 2 1504.38 39.,3** 0.04 Sort x height 1040.48 12 40.87 2.38*': 0.01 Sort X radial position 1861.01 4 465.25 12.16*" 0.02 IIeight x radial position 1003.90 12 lSH.(i(i 4.15** 0.02, Sort x hvight x radial position (j01.(i8 24 25.07 0,6(i 0.00

Hesi<Iual 73,096.90 1911 38.25

Total 83,104.48

earlier pnblication (Arganhright and Wilcox 1969) and is the quotient of weight increase dnr to oil absorption divided by initial oven-dry wciglit. Data were analyzed by analysis of variance using the program NYI3MUL (Anon 1969; Finn 1968) on the U.C. Berkeley CDC 6400 compotcr. Dunn's multiplc comparisoi~ procedure (Kirk 1968) was used to determine the variable levels responsible for significant differcnces; the proportion of the total variation accounted for by each variable (6') (Kirk 1968, p. 134) also was computed.

RESULTS

Mill data Table 1 shows the results of determina-

tion of board-end area. The difference be- tween the total number of boards in the study (1078) and the total samplc size of data in Table 1 is the result of inclrtding boards containing more than one radial position color codc in the data for different radial position categories. Results of statis- tical analysis of these data appear in Table 2. Ilunn's procedure showecl that for the variable radial position, differences between all levels were significant; for sort the dif- ferences between sinker vs. sap and corky vs. sap were significant; while for height, only the differences between height 1 (butt log) and hcights 3, 5, 6, and 7 were signifi- cant.

Thc mcans shown in Table 1 reprcscnt the square inches of board-end area that an average board in each category would be

expected to have (the majority of hoards in the s h ~ d y were 2 x 12 inchcs with a nominal end area of 24 sq inches). Pcrhaps of greater intcrcst are the data for total honrd- cnd area for each category (Table 3) , ob- tained by multiplying each mean value in Table 1 by the number of boards contribut- ing to each category. These data indicate that for this samplc the hulk of the sinker sort came from the pith and middle thirds of tho lowest two logs, particularly the middle third; the hulk of the corky sort came from above the third log and from the pith and middle thirds in nearly equal pro- portions; and the hulk of the sap sort came primarily from above the second log and from thc middle and outer thirds in nearly equal proportions. In cornmcrcial practice thc sort indicates the amount of timc re- quired to dry lumbcr in a kiln, with corky requiring the shortest, sap an intermediate, and sinker the longest amount of time ac- cording to an approximate ratio of 1:2:4. Approximately 45% of the hoard-end area (or hoard volume, assuming a constant length) was sinker, 39% was sap, and 16% was corky; 30%, 4270, and 28% of the board- end area came from the pith, middle, and outer radial thirds of the logs, respectively; and approximately 43% camc from thc low- est two logs, or the first 32 ft of stem above the stump.

Ff~nrlamental properties data In order to determine the effect of 1oc~-

tion in the stem upon wood properties, data were segregated by height and radial posi-

TABLE 3. Toinl board-end orerr (sq inches)

Drrina sort Sinker Corks Sap

Total across Total across Total across all radaal all radial all radial Totnl for

Radial polition Pith hliddle Outer positions Pith \fiddle Outer positions Pith ~ i d d l c Outer position< height

H ~ i ~ h t (16-e IOESI

Total for sort by radial position 2875.1 4188.9 2354.0

Total for sort across all radial positions

Corks

3315.9

Total for radial position acmss all sorts

Pith Middle Ovter

6173.6 8885.3 5856.7 Grand total = 20,915.6

58 W. WAYNE WILCOX AND W. Y. PONG

% i .- e b 7 8 3 - .- - ." . G - .- .B .I. 8 . Y o 6 n

3 6 Eg r e M 3 ". 3

4 Z 3 a 1"

-? ?: '1 7 02 oq o m o 01 a

= . z z E m - ? + : + :

o m o c o m c z - 0 -

- P ; w - 3 " *, - . ," 6 7 G

" 2 .- .. f. 0 C

r-v L O W w w c- m a a n l n ~ ; i s a 0, 3

+(D d m w m m m , - d m - C I-Tr 0 1 3 - 3

b

- - z w.C.Ei ;me

3

- 2-3 $iB 4

o 0

6 C 30 .- d

... s e l . .. . 2

8 8 $ 8 : .;; .;. .;. g .;, U .-

- r - : 4

.,

.u

.i

s - .5

e - c:

c . L V I myi ~ y i ~ y i 3 g

DIC m x t-t- -01 O E 1 3 3 C 1 - a P I 3 3

e m 210 o -in m m 3 PI i 3

01w 0 0 mv) -a o w q- 2 2 m C,

10- ~ 1 3 3 3 m 3 w 01m 03 m a m 0 1 3 El

*I m 1- t- w w >" 01

.- 5

2

.; c J

5 - ."

5 E r - 1 0 m 5 'p L? +:

a $

013 m - n - s a r n - - o ,

$ " Z d 6 S w - M 3 N

0 m lo m n n 2, O N 0, 0 a n

3 m a 8 r l w t n m

C ? ? ? Z m a 0 1 0 c m o n a m 3

c t - w m 0 0 1

z m d m dl? o m v (1 m

m m 0 1 m n C? n

3 m m m n m a a t - W Z W z C 3

m a m a o w o o en, m o ~ m 3 m a * - 3 3

a - m m a m 0 3 w0, a 3 N 3 L O v ) 3 3 - . -

m a at- -1-1 an t-0, 3 C l mrr o w 3 3 3

u

2 7

c E

3 '" r- 6 p.9'c em'g 2..

- : 6

E

- - .- ; g%?g : ~ g

2 - n E L

6 0

2

.- t. C - 2 2

.

.a . c : G e 3

r, - 4 G .- P L 2 : C z - C $ 2

5 2 z ?I z t: '? '1 2 0 C N O N o m

2' 0 * d a d 4 d i - 0 fi N

4 3 9 2 6 m m d g d z d m

2 1 3 3 3 m m

;a- ,F;e , p z - -

1 - 9 m c1 I-

s

. 2 : s .-

- - a 01 3 8 % 0 t- 2 2 +

3- 3

.- z

c E

- - m- 2 a.c.2 51.z e " 8 4 ..

8

- - .; - 0 * B - % d - 0 * C1

n; ' ? m q LO

DI C 0 zN2,zw 2, F1 0 1 v)

z

E E

3 u C- E u.9.P "1t g a 2

* < I _

* - .$ E 5 r

m 2 , 3 1

9 n 9 In

5 l n g a w in a:am 6 3

o n ' ? 9 o ! 3 10% a: m m a ~ 2 7

C " ?'O u? E S n l a w m m n a CI EB LO

3 E "4 6 e...

C? n m 0 . 3 :a mzt-;: U 1 I_ 4 a - " k

(11

'j 2

1 1 0 r

8 z '8 - ,,

3

m (? E! 2 w m o c m m w

m D c u 3 2 <.

S E

C9 1 C4

W I m FIB WOOD I'IiOPERTIES 53

T A ~ I . E 5. Annl!/sis of vnrinnce table for data rc~vcsentcil in Toblc 4

IIuigI~t within test 3,(M1,089.03 16 927,5(iX.06 24(i.l8**' 0.40 Rodinl position within test 1,078,3(i!2.65 16 67,397.67 7?.91"'* 0.14 ITrigl\t x rndial position within t s t 1,489,035.68 3.2 46,532.37 5!1.,M** 0.20

Residual 1,200,788.98 1299 924.39

Total 7,409,2776.34

tion, using the same criteria for these var- iables as was employed in the mill data. Table 4 gives tlir mean valnes for thrsc data. Recar~sr of thr noncomparablr nahlre of data in thc various tests, a ncsted dcsign was rmploycd for the analysis of variance.

TAI~LE (I. Dt,nn'b n~tdtitdc ~om~)ari.sonb for datfl of Tublm 4 mil 5

Sipnifi<.,loar <d 'iifferenccs

Test Moistmc content l v s 2 * n.s.

1 3 *c* *" 2 3 n.5. **

Specific gravity 1 2 I* ** 1 3 *t ** 2 3 n.s o.s.

Tangential tooshness 1 O ** n.s. 1 3 ** n.s. 2 3 II.S. n.s.

Hndinl toughness 1 2 ** n . ~ . 1 3 lid, 11,s.

2 3 n.s. n.s.

T;lngcnti;ll slirinkngc 1 2 n . ~ . n.s. 1 3 ** ** 2 3 * *.

Fcrr h~igb!:, Lurl 1 = I ft, P = 34 It, nnd 3 = il. i t ; for rndial g n s ~ t ~ o n : Irvcl 1 = pith, P = rniddlr, and 3 = <niter.

I,** = rimiril.nnt "I 1% level; = significnnt at 5% Icvcl; n.s. =not sicnificiolt ill 5% Il.vrl.

Results of this analysis are shown in Table 5, and results of application of Dunn's multiple comparison procedure to these data apprar in Table 6. It is appnrcnt irom Tables 4 and 6 that moisture content, spr- cific gravity, taiigcutial and radial tougli- ness, and tangential and radial shrinkage wrre generally highest, wherc diffcrcnccs are sigtiifica~it, in the butt log (drcrcnsing with lieiglit) and in the outcr radial position (decreasing with approach toward tlie pith). One exception to this occurs in the con- sistently high values in the mitldlc radial position of the radial shrinkage data. Longi- hidinal shrinkage tended to be greatest in the outer radial position and decreased with approach toward thr pith, as in the other tests, llut it was lowest in the butt log and increased with height. Absorption was morc! variable but tc,nded to he greatcr in the center two-thirds of the log than in tlic outer third, and to be lowest in tlic butt log and increase witli height.

DISCUSSION AND CONCLUSIONS

It must he remcmhcrcd that these data apply only to this samplr. The accuracy of any intcrl>olation to whitc fir in gcncral clepends upon the drgrec to which the sam- ple is reprrscntative of white fir trees. With this in mind, however, it is usrful to drrivc some hypotheses from thrse data witli regard to white fir lumber.

The data closely agree with those of an rarlirr sample (Wilcox 1968), cxcrpt for showing a liiglicr mean spc.cilic gravity of wood in the present sample concomitant with greatcr toughness and transvrrse shrinkage. A consistent increase in specific gravity with distance from tlie pith was also

54 W. WAYNE WILCOX AND W. Y. PONG

observecl in the present sample; this was not recorded in tlie previous mmplc.

The values for ti:' in Table 2 indicatr that little of the variohility in mill data (Table 1) is accounted for by thc selcctcd indcpen- dent variahlrs. JIo\vrvrr, in the fundamrntal propvrtics data (Tablcs 4 and 5 ) the two indcpriident variablc~s account for large por- tions of the variability with height the rliost important factor, accounting for nearly 50%.

With regard to mivtures within the same board of wood which normally would be placed in two different sorts, it is apparent from Tablc 3 that such mixtures must in- volve primarily sap-corky combinations aris- ing from upper logs. This follows from the fact that both sorts appear to comc primar- ily from above thc sccond or third logs, and both rrly heavily upon material from the middle one-third of the log radius. Unfor- tunatcxly, these are also the regions in the tree having the greatest differentials in fresh moisture content (Table 4). As upper logs also tend to I,e smallcr in diameter than lower logs, any board coming from this region would have a greater tendency to include material from all radial positions than would boards from lower logs. Mixed I~oards coming from upper logs would tend to h r included in the sap sort (which com- priscs approxin?atcly 40% of the Imard volumc) brcausc., for drying purposrs, thcy would have to be treated for thc higlicst moistttre contcmt present to prevent thcm from I~eing inadequately dricd hy the end of the treatment.

Over 40% of the total volume of boards in tliis study came from the lowest two logs. According to our data, boards from these logs should have the highest specific gravity and the narrowest range of moisture content of any in the tree (Table 4) . This would snggest that boards of tlie highest quality could result from these logs, aud that proh- lems in drying associated with large mois- turo diff(~rcntia1s within thc samc hoard could I,(, minimal, lad ing to a greater con- tinuity of board quality during each step in processing. Our data show that more than 80% of the board volume in the lowest two logs is of the sinker sort. These logs are also the portion of the tree that has prop-

erties most typical of wetwood (Wilcox 1968). If what we have suggested is correct, this would indicate that the presmcc of wetwood has little or no adverse rffect on quality of white fir. Furthermore, this suggestion placcs greater valuc on thr lowest two logs in whit(? fir, hrcause thrsc logs appeilr potentially to provide not only the highest quality boards but also ncarly half the total volunle of boards in the tree. Tlir relationship 1)ctwrm stem location, drying sort, lumber grade, and degrade will bc explored in another papcr.

Some of the boards falling into the sap sort also comc: from lowcr logs, and would share the properties discussed above. How- ever, because the sap sort must also contilin many of tlie mixed sort boards, which may bc responsible for some procrssing dif- ficulties and degrade, it might be advan- tagcous to make an additional segregation into which mixed boards could 6e sorted. This would make the propcrtics of tllc rcsidual sap sort more uniform and would protect potentially hig11-cluality sap boards cut from the lowcr logs. It would also rr- move boards that may cause drying proh- lems to a separate, smaller group where it would be possihle to dry thcm on a diffcr- cnt schedul~~. This could rcd~lce dcglxdc: in this group of l~oards if, it1 fact, this is n problem. Data dealing with thesc hypoth- eses also will be considered in the paper dcaling with ltlmhcr quality and degrade.

REFERENCES

Ann, P. E. 19fifi. Defrrt estimation for grand fir, Engelmann spruce, Douglns-fir and westcrn larch in the Blue Irlom~tnins of Oregou and Washington. U.S. Forest Spw., Pacifi~. Nrrrth- west 120rcst Range Ilnp. St;!., Portland.

ANON. 1969. NYBMUL. Univnrintc m ~ d rnolti- variatc analysis of vnritmcc nntl cnvnrinrrcc. Comp~~ting Crntrr P~.css, Statr. Ilniv. NPU~ York, B~dfalo. 70 p.

ARGANIIIII~.I~T, 1). C. , AWE W. W. Wncox. 1969. Comparison of parameters for prt,clicting pcr- meahility of whitc fir. l'roc. Am. Woorl- l'rt~scrvcrs' Assoc., 65: 57-62,

FINN, J. 1). 10l??. Mull iv~~riarl~~ . . . univnriatr ;md moltivnrinte nnnlysis of varinncc, covar- iancc, ant1 rugrcvssion. A I'ORTRAN IV pro- gram, Version 4. State Univ. Nrw York, Rt~f- falo, Mimrogmph, 108 p.

~IAATI~RY, C., R . W. DAVIIISON, Ahn B. S. CTIAU-

WHITI< Fnl WOOD l'IIOI'ITI1TI~S 55

L lSJG1. Wehvoocl, bacteria, anrl in- creased ~III in trecs. U.S. R,rcst Serv., Forest l'md. I.ab. Rrpt. No. 2215.

]AI:K%>N, G. [I., J. \V. ~IENLT~Y, ANIIW. L. JACKSON. 19&3. Lop diagraming guide for wpstprn soft- wonds. U.S. Forest Scrv., P;wific Northw<:st Powst Range Exp. Sta., I'ortland.

K ~ K , H. E. 1868. Experimental desipl: pmce- ~ I I T P S for the behavioral sciences. Brooks!Cole I'ublishing Ca., Behnont, Calif.

I'ONC, W. Y. 1871. (:r;nlc ant1 volume change of C;difornin white fir Imnhrr frwl rough grcPn to sr~rfacr dry. U.S. Forrst Sen,., Pacific Northwest Forest Ri~nge Exp. Sta., Portland ( in proc<.ss).

-, R. h4. BASS, AN^ H. D. CLAXTON. 1970. An automatic phntwlerkic triggering rnrrh-

anisrn for a dntn-recording campra. U. S. For& Srrv., Res. Note PNW-122.

-, AND 6. 11. J a c r s o ~ . 1971. Diagraming true fir logs. Stlpplcmant to westcm s,rftwood guiclc. U.S. Fon,st S c l ~ . , Pacific Nnrthwrst Iilrwst Rang? Eap. Sta., Portland (in procrss).

S ~ I I T ~ , 11. IT., AND 1. R. ITM MAN. 1%%. The segrpgation of whit<, fir for kihl drying. IT. S. Fnrrst Serv., Pacific So~~thu,rst Forpst Range Exp. Sta. Hes. Note 167. - AND - . 196Ob. Drying rate of white

fir hy sol[rt!gations. L1.S. Farcst S(w., Pacific So~~thrnrst Fmcst Rango Esp. Sta. Rrs. Notc 168.

WTI.DOX, W. W. 1868. S O I ~ C physienl ant1 m e chanieal proporti~s of wrhvood in whitr fir. Forest Prml. J., 18( 12): 27-31,