Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
-- 6 Erh; bit _ PROPOSED COt.iPIL~:ON OF LIZ}; DATE OF S;EDI~TTS W. C. ~nbein University of Chicago, Chicago, Illinois Prelirun~ar~r ~n~x~coment of the Sub-Cor~.aittee for the Compilation of Statistical Data on Sediments Tntroducti on ~_ The Sub-Comn~ittoo for Lee Co~pilaho:n cf Statistical Into of Sed~- ments was appointed by Park or D. Tras}<, Chin of the Comrnittes on Sedimontation of the National Pcso~rch Council, for tho purpose of making available in convenicat form the qu~titc.tive do.ta on sed~mc~ts which in now widely scatt¢rod through the geologic literature. personnel of the Sub-Committee is as follows: W. M. Cogen, Shell OF Comply, Houston, Te:~s W. O. ~7(rumbe~n, University of Chicago, Chairman G. H. Otto, Soil Conservation Service, Pasadena' A. C. Tester, F~iv~sity of Iowa The present Calif Ortega The Sub-Co~ttee has not met as ~ body, but the Printer has conferred with Trask and With individual members of the sub-comm~ttee rogard~ng some of the problems ~,rol~red In the co~ttea's work. ~ a result of these conforonces it was docidod to start the work of tho committee on mLoch.~?~.~1 analysis data, both because of the large number of such c.nalyses av=~1~' arm becmso of the grouter ~ed~to USE which may be made of s' ze data in the corupax~so:~ and interpretation of sent. Several practical difficulties arise even here, however, in the mere task of oc~mpiling the data. Some analyses are Comply, due usually to techniques which do not include the finer sizes; other o:nalys¢s were made with so few class contorts that inadequato crate arc a~ra~lablo for significant sumr~ies; and some papers include over ~1 graphs or condensed Barrios of statistical Bruce, enrich lordly precut further analysis of the original size distribution. In wow of the unsatist`~tory state of many earlier Ed some contemporary Ales, it socked ad~ris~ble to issue a prey state- mcnt indicating the ,:E`~:ro ~d scope of the ocmm~tto~'s Work, 'as wo11 as to suggest more suitable procod-~res for th& publication of sedimontar~r data, to avoid d~ticult~ as in the Rapture. interested readers an opportunity of expressing their Mews about the work, as well as to pert the =~-co~ittee itself to crystallize its Ah? ~ report W~ 1~ als o =~ford
Exhibit ~ aims and metho d o ~ at track . Scone and Purpose cat ~ Co:.nmittee's Qu~titat~e intonation on sediments includes ~ wide variety of analytical data. NQt 00~; size distributions Qua sediments, but shape and roundness distributions, mineralog~c~ studies, pet,~otabric Saudi es, and the investigation of such mass properties as porosity ard pe~cabil~ty, are frequently encauntcrod in the litera~ro. It is ob~ou sly beyond the soopc of any single c~nittoo to attest a compilation of this vast field. The present committee is cornmi rosined with the Glucose of co~'l- ing size data pr~r~ly, with ~ possible e=ens~or~ to other characteristics such as shape and roundness. Parallel work on other aspects of sediments,, especially heavy ~Li:aerals, may well ocei;py a similar committee it the present venture i s rmso~bly successful. . . . . . . . Sedirae~tary petrologists who- wish to c~pare their f~nd~x~gs. =~ the results of earlier studies arc forced to search c,= over-i:acroasing l~teraturo. Goneral goolo~sts Ache wish to gain a co~reLcnsi~c p3CtUrO of the characteristics of any group of sediments bind few quantitative data in tows or rofor¢nce books. ~_. ~. ~ An cutout to offset this d~¢ulty was made in 1931 by Wont~orth, v~hc published nearly a thousand histc'~;raTn.~ of common sodimonts. This oxcollont sundry is widely used, a~ Ingests the desirability of preparing ~ fairly complete compilation of ~11 available data. ~o NATO task of compute on Is prodigious, inasmuch as shore must be thousands of mechanical analyses in print. Such an undertak~ir~ can only be assumed by co-operative Short, and the present ccqr~iaittee has for its purpose at least aL.preliminary e~lorat~on at the field. ~ The speci:~c purpose of the co~ittee's work is to assemble into convenient fc~ the availab] e data on mochameal analysis, so that workers may have an easily accessible summary of the observed characters sties of sed~m~ts. In the.decade which followed Wentworth's work shore has been increased emphasis on the statistics troatmondt of tnech~ic~ analysis data. This wider recognition of tho goonotrica1 meaning of statistics parameters, =.nd their greater use in the intax~prct~t'-n of the data, suggests that the co~ttoe's su~=y may co=~niontly be published as statistical data rather than is ~st<3gr~ or cumulativo cunos. involved in two cost of publication a3Lso.fca~c~r this solution. :E?actors Present plans are accordingly to a=?~nge sed=~n=,ts into various croup a. such as beach sat ~ dune sand" river gravel , etc., and to compute the statistical measures of each sample. These data will be summarized in tables, giving ~11 information regarding; the source of the original data. For those sedimentary types in Mice; su:eficiont analyses are available, the tabular data will be ~rther strived in teems of the general statist3;ca1 cha:mcterist' as of the sects. . Thus any work cr who wishes to compare his own data on pa:rticul~r sot with previous Priorly may }.eadily do so by noting whether tho general characteristics of hi ~ ~, ~, ~_ _ a,
Exhibit A .,. _ supers Ogre with tho lo~r cavorages obtained from tho overlies data. Similarly, the characteristics of ancient sediments may be compared with the data given in the tables to see whether the unknown samples fit Alto any recognizable geIl~a1 group. In order to facilitate groping the sediments, it lo preferable to confine the data in the tables to xc3cont sediments collected from know on~rom~nonts. The ~nclus~c,~ of ancient sediments Which mar have been interproted as belonging to particular genetic titles woul d introduce inferential data into what should be a strictly facial picture. It is generally recognized that wide variations hong sediments from the sane environment render it doubtful whether in the average case it =11 to possible to identify unknots sediments from their size charactoristics aloe. It is partly for this reason that the corm, ttee hopes ultimately to extend its work to other properties o* sediments. It is not unlikely that if Qu~t~tat~ve data on several other ch~.racter~stics of seas Arch as shape a~ roundness) were av=1 able in summarized form, ~ given sediment might heave characte:rist~cs common to several types in terms of size, a~ yot be di~erontiated in respect to some other characteristic. -4mong Widely dx~orent -sod1montary- rows it is possible that size chare~te:rist~cs along may be helpful In Pontification. Chow ~ of Stat! sti Cal T`teasures . . . . . Rocent increased emphasis on statistical state of moch~ic~l analyst delta carries witch it ct roast ho disadvantages. ~e first is that workors uso ~ ;~ e Amity of statistica1 norametors, 50 that direct comparison With the Gore of others is frequently difficult. l.~Q~eov=, there is ~ ~ncreas~x~'g tendency to publish only the statistical ~x~naries, rather than the complete analytical data, so that oth~ workers cannot recompute the data into other measures. One solution which suggests itself ~ s that all workers adopt the same statistical dewees, but such a solution Is probably premature. Experience in other scimees which use statistical methods has $~0~ tot 000 cannot predict, ~ pr~ori' which at several stat' stical approaches is the most significant for a given set of aeta. IJntil analytic al theories of the factors contrdlli~ the statistical parameters of sed~:acnts are developed, it seers safer to use several approaches. Sorae choice must be mad o among the many statistical approaches, however, to avoid an in:fin~t~ set of p~;~m¢~rs Ace each sample. Within the past decade stati$tica:1L methods based on co~vontiona1 met hermetical theory have tondcd to supc~sede mp~r~cal or unconditional methods, but such specks methods are still rathe3: widely used by European workers. Altho~g:h no prediction ~ r be made of the exact form of the most s~ti- cant parameters, e~erimce in other bells has ~;o,~n that those'`neasures whim integrate Dim the great body of probability theory are generally
Inhibit h ~ _ __ ~ __ most ~=ccess~ll:r applied to analysis. Among such measles are the quartile and moment measures, expressed either in d~mneter terms directly, . . . .. . . or as their corresponding geometric and logar~tb~nic transformations. In addition, certain positions values, such as the modal diameters, axe useful amok distributions ~chhave more then ore ''max~mum,'t In general it may be said that sedimentary experience has shown that geometric and ~ oga-rit1~m; c misuses are more useful ~n interpreting sediments thy arithmetic mea=res, t,ecause tho former lair the size tractors in m~ measures as sorting and duress. Arthur, the geometric am logarithmic measures may be Preformed comparatively easily, so . . . that both are readily available for comparative p~OSQSe Peso factors suggest that the data of sediments be summarized in tams of geometric and aog,~rithmic measures expressed in both the quartile and moment notation. This approach incIndas specifically the following measures: 1 ~ Median diameter An mm. Lois averse grain size represents the middlemost diameter of the distribution. . . . . . . . 2) Geometric mean d~neter in ran. lhis average size iB it's antilog of the ce=~of gravity (first moment) of the Sari thmi c di stribution-. 3) Modal di~eter ~n mm. . The modo represents the highest point of the frequency curve. TYhm several modes are present they may be rm::re si~it~cant for summarizing the di stribu~on then aver single mean bad up. 4) "Sorting coefficient, " Go, (flask) of the so . Tha s is a nea sure of flue geometric spread of the c ontraJ half of the d' stribut~on. 5) me logarithmic standard donation' 0~' This is based on the second Comet of the logarithmic distribution, and it measures the average spread of the distribution. 6) Tho quartile ~kow:}ess (Trask) of the sediment. The log of this measure to the baso 10 e~esscs the asy~etr~r of the central :~f of the distribution. 7) The logarithmic skewness of the sediment, Ski, This is based on the third moment of the logarithmic distribution , and it measles the asymmetry of the distribution. Detai ls of the measures: ~ described above, and suitabI e m ethods for computing them may be found in Trask (1932) ~ Erumbein (1936), and l;tumbein ~ Petti30hn (1938' chapters 8 am 9~. The graphic deto=aination of moment measures is described by Otto (19593.
- - lo - - Tllu$tra~ve E~ples Stat~stica:L Sullies ~ib~t A _ It Is believed that the following data are essential for the tabulated summaries: 1) A committee retere3ace number for all sappiest Reich will serve in subsequent quotations by wetters who use the data. ~, 23 Ihe source of the data, with the original autI~or's reference numb er . 5) Lee geographic location of the sample. 4) Me nat=e of the sect. 5) The completeness of the original data' in tens of the Thumbs of size grades used in the analysis, a~ whether or not a full report of the analysis is given by the author. 6) The method of analysis used. 7) Me several statistical values, together with some indica- t~on whether they wore given by flee original author or Were computed by the sub-co~ittee, Scruples of tables whith she' this ~ntorr~ation are included here to illustrate the method of presentation. O my, lye s ~ exam? es are t ent ative once due suo-cor.~ee urges ~nrerestod readers to offer criticisms and suggestions which will ~ the compilations more U$~] . For the examples three kinds of sediments wero chosen, largely because the data were a~r~lable in files at i;~¢ University of Chic~o. Table ED-T shows a number of beach suds, table PS-T includes Alluvia sands, and table GT-I covers ~ number of Glacial tills. Etch table has an index number, end within each table the samples are numbered serially. Thus for reference purposes one May cite the table and sample numbers. Me principal items of intonation are given in separate columns, and under "Mark" are symbols which-Wish data on the less ir~nediately- used information. ~ bibliography of source material accompanies the tables; a list of the symbc>Is used Is given below.
Exhibit A , _ a) En hi ~C:- En o En ¢ '_ ¢ (Q ~_' H En cO En H En O V c! PA ·e as _ _ ·rl &] -- 11 -- I S=-4 I . H U) o [a ~ CD ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ . - . - . - · - · - a~ ·^ a~ .^ . - . - .% ·^ a. · - . cl? C3~ C37 ~ C9'C37 Cal ~ C37 Cal ff' ala C31? 1:37 Cal =~=Q~=Q~ m~= a- e~ e~ c- ·- a~ a~ e~ c~ ·^ · - ·^ ·^ e. e~ C~ ~ ~ E~ ~ C~ ~ ~ ~ - - cn ~ ~ ~ cn ~ ~Q di , ~ ~ ~ ~ c~ c~ ~ co ~ cn ~ ~ ~ ~ ~1 --~ ~ ~ ~ L=~-~ 0D ~ C~2C ~ o bO 0= ~ V: ~ t_ _ I.` 1- ~ 1 o 1_ . i ~ ~ - G - . o~ ~ ~._ i . 1 ~ o ~rl CD o CO 1 .. :- o r ~C;S ~ ~ di c~ GO O ~ ~ a) * C~ ~ di d~ ~ O ~ ~ ~ ~ ~ O O O ~ c~ o o o o ~ o o o o o o o o o o o CD Cn ~ C~ ~ O ~ O ~ ~ ~ 00 U) (D ~· ~e ~ O ~ O ~ ~ t9 C~ ~ y3 C ~ C~ ~ C~ · ~ · ~ ~ ~ I. ~ - ~ ~ ~ ~ ~ .~e ~ ~ ~dt - - di d4 - did~ u: ~ ~ ~ ~ Lo ~ ~ o ~ ~ ~ ~ cD ~ c~ ~ ~ ~ ~ c~ ~ c~ ~ o ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 1 C~ cr~ ~ ~ ~ 03 0 ~ ~ ~ ~ 00 00 ~ c~, C~ ~ ~ ~ ~ ~ ~ C~ 2 C~2 C~ C~2 o ~· ~ c~ di C~ ~ ~ ~ O ~ a:) ~ 0 ~ ~ -- o) Cx2 ~ ~ ~ ~ ~ C~ C>2 C~ C~2 C~ ~ ~ ~· · · - o o o o o o ~ l ·. · ·~- ~ al~ ·- ·- ·- ~,-rd ~ ~. C:) O ~ O O O H H H ~ H H H H , LS) ~ ~ CO ~ O [-~ .~l ~1 .~1 C~ cd ~ V ~ P:l ~ ct ~ V 1 ~ 1 ~ 1 1 ~ ~ ~ =' ~ ~ ~ ~ ~ c ~cx2 c~ c~ I I I t I I t I I r~ ~ ~ d~ ~ ~ ~ ~ ~ ~ -~ di t ~ I I ~ ~ tO I~ p:1 =., p:t P~ 1 ~ 1 ~ ~ 1 1 1 1 ~ ^- ~ ~ ~ ~ ~ ~ h ~ ~ h P~ ~ ~ ~ ~ ~ ~:d 5 ~ ~ ~ V V ~ V ~ ~ ~ V () i^ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ O C$2 ~ ~ ~ <o ~ ~ ~ O ~1 ~ ~1 ~1 ·rl h ~4 I· ~ ~ o C~ ~ o - i ~- a) : ~ ~ v) . tD O ·rl o ~ o v ·H c) c) s~ o o oD ~rt 0 ·. D 5- p ~ C~2 .H ~ o ~ ~ ~ 0 1 di ~ o o c~ s~ ~ G) c: p4 o p a' ~ · ·~ ~ v `m m ~ - o ~ ~1 j;7*~' G> ·rl ~Q ~ o ~ s~ 0 a) .~.rl er: v . ~ co ·~H ~ o ~ ~ ~ ~p~ ~ ·. c~ ~· p ~ ·~-.` l cO v ~ a p~
1 o I rQ H En C H En En Exhibit ~ ~ - - ~ · - ~ ~ s. .. ~ . - e~ · - .. . - e. C. ew ~ Hi Q ~ ~ ~ 1~ ~ ~= £\ ~ A ~ . - a. ~ em . - c - e - s~ . - 0O ~ CO ~ ~ ~ ~ GO Car CO CX) ~ ~ ___ CO 1 ~ o ~ ~ ~ 0 ~ ~ ~ ~ cut ~ ~o o o o to o to to ~ o ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ , ~= E ~ I I I I ~ ~ cn e~o ~- <~ ~ COLQ~ ~-LQ--a~ co ~m u: O ~ ~ C~ ~ C~ ~ ~ C~ ~ O.C\2 ~ · ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ C`2 1 1 1 <~ - .-1 CQ : ~ a) 1 ~ ~ - ~ ~ ~ 1 l ~ ~ 1 ~ ~ o ~ ~ ~ ~ a' o ~ , ~A _ ' I '--~ · ~ 1 l 1_ ~ 1 CO o ·~ j c~ cn tD ~ ~ ~ co ~ a~ ~ 0 ~ (D ~ o ~ ~ CD ~ ~ CD C`2 C~ C~ LO ~ ~ ~ ~J o o ~ ~ ~ ~ ~ o ~ ~ o ~ ~ ~ ~_ C~ U) ~ ~ ~ C~ C`2 ~ ~ ~ ~ ~ C~ ~ ~ ~ ~ C~ CO C~ C~ ~ ~ CD CO C~ . . ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ _ . C~ o L~ d ~. di ~LS~ di O . o . LO o . ~ ~ ~ ~ ~ ~ o ~ ~ ~ ~ ~ 1 di o ~ ~ ~ o ~ ~ o o CD ~ oo ~ C~ =, o ~ ~ ~ C~ C~ Cx2 r~ Cx2 ~ · · - · 0 .e · · · · · · ~ (D ~2 ~ ~ ~ c~ cr~ ~ ~ ~ ~ co L~ C~ C~ O ~ ~ ~ ~ ~ C~ ~ ~ C\2 · ~ ~ ~ ~ o ~ ~ a | ~r/ ' ~@! ~ ~; ~ r ~ a ~Cl ~O O ~ O O C~ ~l~ C] O ~ ~ H ~ ~ H ~ ~ ~ ~ H a) ~ ,- ~ ~. . . C) ~LO ~ ~ ~ ~ ~ ~ ~ O ~ ~ff ~ ~ U: ~=) CD CO (D 00 0~) OC~ OC) ~ a: co ~ cn 1- 1 1 1 ~ 1 ~ ~ I t ~ t t o ~cP ~ di ~ '~ ~ ~H ~ C] ~ ~ =H rS: 1 ~I I I i i. 1 . I t I ~ I . =. - .4 . P.~4 H l _~ C) ·H U) C~ C~ O ·rl ~4 ·,1 co 1 0 LO Q4 U) o , ~ P~ . P4 C) · - C) ={ C) O · I 1 i ~ ~ ,_l, · ~ ~ O a) ~ ~ e P ~ ·-4 C) a) 5- c~ ~ ~ ~ =- ~ ao cn ~ ~ c~ ~
Exhibit ~ _ H 1 a. C) a) Sat 0 a' V: H A o ¢ H En to H EN U3 H .- .:q · - a' -_ 13 -- GO V' owe - o V) . . cat in ~ di O ~ ~ ~ Cat o o ~ 0 0 0 cat di di · ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ co hi ~ ~ ~ ~ ~ ~ lo, ~ ~ ~ ~ ~ (~ ~ ~ t;- ~ =, · N ~ can ~ as ~ a: car ~ ~ ~ ~ ~ cn O d' ' · · ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ - ~ ~ ~ di di ~ ='di ~ CO Cry Co CO dot ~} 'I C~2 C<) ' h 0 in' red ~ ~ .~ ~ _- . ._.... O ~ ·' c) a, == _. . a) ~ ' C) As C) o ' =~$~QaQ4~iP4~4~1 ~4~P~ CO (a ~ ~ ~ U) U Cal ~ U] Ci:) Gil . - cm, .^ a~ am .^ . - act . - e. .^ ec, e., . - a. cec3~=cic3ra~ Gil CACTI 0 0'0 0 0 0 0 0 0 C),O 0 0 0 O. ·- e~ e~ c" e~ ~ es~ ·- e~, ·~. .~' .~., e- ee, ·-' . ~f ~ ~ ~ t~ rH ~ I~ ~ t~ r~ ~-,-~ ~t ~ ri r~l ~ ~J .~1 r.1 ~ - 1 ~I rd - 1 - 1 ~1 ~ ~ ~ ~'~ ~ ~ ~'~'~ ~' ~ ~ ~ I 'q' l C~ C~ - ' - O - 0 ~ ~ O · ~ ~ ~ ~ · · -, e~ a~ ah e~ -. ''*~ e ~I Cx2 ~l ~1 ~t rd d.' c~ - 1 ' - - 1 ~t ~ C~ r~l rH ~ r~ ~ =1-'rH O O O O O O O O O O ~ O O O O cD ~ ~ ~ ~ c~ c~ ~ c~ ~ o c~ ~ cO o o o ~ o ~ ~ c~ c~ co ~ ~ ~ ~ ~ ~ ~ o ~ ~ o o. o o o ~ o o c;; o o O ~ ~ C' ~ Co di ~ ~ (D O ~ ~ cr, 00 0 0 ~ ~ C~ cs2 Codi di.---O o o o o o o o o o o o o o o o · ~ ~ ~ ~ ~ ~ ~ ~ ~ · ~ ~ ~ ~ H H H H H H H H H H' - l H H H ~ V) S~ O O ~ C~ ~ ~ - ~ CS) ~ ~ ~ ~ ~ ~ ~ {b PA.. I ~ l, =. ~ ~ ~ ~ ~ ~;) = M CO C~ ~ ~ ~ CO I I ~ I ~ I I I t 1- 1 ~ I ~ I S~ S~ ~ ~ ~ S~ S~ ~ ~ h ~ h ~ ~ ~ ~4 ~ ~ K~ ~ ~. ~ ~ ~ ~, - V=~V~V~ ~G -,_ ~ L~ ~, r=7~ ~L~ ~ ~ ~ ~;~ t~ t;~ 1; ~ >; ~ . . ~ O p:; ~i ! 1 · ~ c~ ~ ~ ~ c~ ~ ao cs: 0 ~ C~2 . · ,- ~,- .' ·H v bf . ~ . o ·rl . ·~ U . ~rt SD'= o o ~ .o -. 0 d~ .e ~ O ~ p · ~ i2.- ~ . r.-' a' o St c) ~rl ~ e I S~ O *e ' Itt) '' V O ' :: . C) a) ~'
--- 14 -- bible , ~_ As extensions of each table are published, the serial numbers (column 13 mill be merely continued under each tabular head. The source of the data, given in colors 2, include the initials and first two letters of the authorts ~e, the it - r of publication, and the author's original sample number. Thus J=a-14-65 refers to sample 65 of a paper by Udden In 1914. The full reference is given In an alphabetical bibliography, arranged according to the key letters themselves. The third column of the tabs e indicates the location of the sample. In the United States it is expressed by state; elsewhere by country. Condemns 4 to 10 list the statistical parameters of the samples. Because of the wide variation in methods of analysis, am particularly in the n~a~bsr of size grades used' it is necessary to distinguish among the parameters in te~s of the w reliability as numerical data. ~ beach sand analyzed with three silo grades cmnot Gove ~ close an approbation to the true paromoters as the same sand analyzed in seven or eight ambler size grades. In order to include such analyses, however, the following arbitrary plan is proposed: a) When the number of size grades used is three, the data will be given ~Q the first signi~c~t file only. b) When four to six size grades are used, the data will genoral- ly be given to two significant figurer. c) ~n seven or more size grades ~Q USED the data mill generally be given to three sig~lticant figures. a) The data Fill never be given to more than three decimal pla ces. me te~ "size grade" as used here refers to the anthor's original data, whether it is expressed in ~ex~tworth Grades, Comrades, or any other system, and regardless of the frecuency in anger grade. The quartile data may be read from data based on irregular ct ass intervals, but th e moment notati on i a; facilitated by regular ,~comotri c interval ~ in the size grades. 3uro~u of Soils grades, for ex~plo, wound have to be reconverted into an equivalent number of YVer:~worth or ]'2-grades, depending upon the ori~;ina1 number of grades reported. At present it sews unlikely chat analyses reported merely as "sand, silt, arm clay" can be adapted to statistical computations or graphed. this may intcrfore wash the use of some a=1ysos b`ased on tho hydrometer method, for example. The above -ply is subject to the tollowi~ restrictions:
All; bit A . _ - 15 - When the data have "open ends I' (i.e., when an appreciable p=t of the distribution is not included in the analysis, as mom t~ne-grained sediments) the number of significant figures is generally reduced by one. In such instances also the skewness values will generally not be included because of Thor fictitious value in such cases. When the data are polynodal (i.e., in&Ye more than one maxims In the distributions the sign: ficant figures arc general- ly xeducod by one, and tho Lowness values are not computed. In connection mththemodos, the tO1iQ=~= p~codurc is Accosted: Tho principal rude is dctinod as tic geometric mid-point of that class harm the largest percontage frequency. Secondary Rhodes arc considered sigr~hcant if they arc prcs¢nt among classes having ~o per cont or more of tho distribute on. Open onds of distributions ~o not considered to bo nodos. Ihe order of listing the Diodes Is the principal r.~od¢ first and secondary Modes in order of frequcacy In the modal classes. Son c secondary modes may occur In the open ends, but these cannot be detected in such cases. The "Remarks" column includes a number of it's which may prove of value for reference Osmoses. These symbol s are arranged in their positional groups separated by semicolons. The following list indicates the symbols used in the current examples: The first position in the column indicates the grade scale used, and the number of grades, where that ~nfox~ation is giver. For example, the symbol W-7 means that seven Wer~two~rth grades were used. The symbols for flee grade scales are: . . . A - Atterberg scale B _ Bureau of Soils scale :B - Mesh scale (onginooring practice) - Forth root scale I - Other scale ~ _ Square root scale W _ Wentworth scale reported. The nm~be~followir~ the grace scale symbol is the number of grades The second position in the column indicates the con:plote- ness of the reported data. Here arc i no? udod the following symbols:
-- 16 -- c f H o Q author gives cumulative curios author Eves complete finale cal data author gives frequency curves author gives his togr~s author gives moment data author gives nc, analytical details author's data includes open ends author gives quartin ~ data Exhibit ~ The third position in the column indicates the method of analysis used. The symbols are: ~ - elutriation h _ hydrometer m = microscopic method o = other meted based on Oden theory p = pipette s = scat Classification of Sediments a.. One of the p~robleans befom the committee is the choice of a Citable classification of sediments which may be used as ~ basis for the tabular subparts. A detailed class~ticat~on would sewn desirable so chat shades of distinction may be-drawn between sediments, but practically this is impossible at ~resent. Many authors describe their sediments in broad environmental terms. and too bide a distinction. not based on sound T.t scat s~cst as ~ first approach to use a broad classification based .~r~mar~ly on gonesi`$ and socond~ar~ly on Bean size. X~nhofo] (1939, p. 51) gives the following classification of on~nrorments: theory, mail lead to over-~Phas~s of minor differences. . Conti nonfatal Torrestri=1 Des ort Glacial Aqueous Pluv~a~ Pi odr~ont V£~11oy flat P3;Luda1 L=~e sweeps Rivers swamps Flat-land 5W0I2p 5 tie sweeps X~nhof o3
=hibit A -- 17 ~- cu:;tr~e Prosh Salt Sp clears ~ cave ~ taxed co:~tine:atal and unrig Li ttoral 3)e] ta Marginal lagoon Study Marine Neriti c ~thyal Abyssal Twenhofel's classification may be considered at ultimate end in the c~pilaticn of $edim~tary data, but for present purposes it seems neces- sary to condense the classification somewhat because of the l=ge amount of data which may not fit Erect sely into the individual groups, owing either to incomplete data from the authors, or because the samples are classifi ed on some other basis. As ~ first approach, th~afo:re, the f o l ~ on ng br eakdown i s su gg e st ad: Wind deposits Dano sand Loes ~ Wired lays Glacial deposits Glacial ti 11 Glacial ollfwash Fluwal sediment s Channel deposits Flood plain deposits Swamp deposits (fresh water) lacustrine deposits Fresh filter lakes S`~]Lt la: Shore deposits (beaches without distinction) Delta deposits
-- 18 ~- . . . . . . [agc~onal deposits? including b.ays and estuaries ~ · ,. . Iv~rino Bottom deposits . . Moderate depth (~s than 100 fathoms) Deep water (greater than 100 fathoms) Icon by ~ ~ I . - _. . . · · . . . The list roughly parallels Twenhotel's classification, although some of the distinctions are less sharp. :Efor exam., beaches usually consist of sand or coarser material, and as a first appro:~:i~tion it seems safe to merge the deposits of lake and sea beached Tt sems unlikely that the difference :between salt a~ fresh water would mate:~ially affect the physical behavior of sand~or pebbles along ~ beach. It is not sate to Wage lake and marine Mud, however, because of the possible coagulating oft'ect of.~ons.~n sm water.. Sir.lilar1y, outrun met elicit Smalley train and outwash plain deposits) are ~ separate group because Virile fundamentally they axe;wnning water ds~?osits, sorae wor1i.<ers may prefer to treat then separately from fluvial materials .,. ., ; As work progresses it ~11 undoubtedly be found tat the classific~l- tion will require n~ditication. Readers are invited to su~;g,cst more appropriate g:~oupi~s. . . ~ . Within each of the procodi~ groups ~ =rther subdivision is made on the basis of mean size. For this purpose it is believed that the center of brevity of; to geometric distribution is the ~prop~atd ncasurc to 1lSe il1O~UCH I it affords a control point who ~ is useful in the physi Cal aiysis of tractors attec.ti~ transportation (Rouse, 1939). The distinc- tion among the sedimonts of Such group hill accordingly be made on the [0110WiDg basis: , . , . -, . . . . a) Geometric r.l~ di;emota~ grmtor than 2,00 ~. - ~V~ t) Geometric mep~ diamotor between 2~00 and 0~065 An. ~ SAND . . . C) GOO~QTriC mOnD. diameter bO11;;~GOn O.062 and .004 ma. _ SI:LT S) GO~iC mmn diameter smullo:r than .004 Ad. - Cam ~ Or cOm~?ilea Data to Practical Work _ . . . . : Bomb tises which may be mean or the compiled; data are the followed;: 1) Alien Rough analyses (at least lOC)) are compiled on a given sedi- m-ent, it wi11 be possible to obtain avera== of the several parameters, so that one may say that the average size of dune sands in general is such Ad =~h ~ magnitude. Thus one may have general magnitudes fob the size, Noggin, am sameness of various sediments as a 'owe. - .. · . . · . · · · . . . . 2) As the data accumulate, i t $ril l be possib le to prearm frequency - distributions of the averages and other measures, to observe how
Exhibit ~ - -- 19 -- indi~du-1 sappy es of a ~n sod~=t vary about their okra mums. For example, it seens likely that the curve of several hundred dune sands will yield a distribution of means having a lesser spread than the some number of outwash sands. The relative homogeneity of various sedim~ta~r Ados cold thus be brought out. . 3) A worker with ancient sadists may wish to dotermirle whether one of his samples may be interpreted as a dune mad. By comparers : the parameters of his scruple with the group averages' it will be possible to detente whether the ~3ivea~: simple lies within the rave of values solo by the group. Under some conditions it my be.possible to apply probability theory directly to this prob fem. ·4) It Sufficient data on other segmentary characteristics (such as shape, roundness, mineral content, and the like) are obtained, it may be possible ~n some instances to identify ancient sodim=ts in terms of the s=aultancous agreement among several sets of paramotcrs, prodding any one parameter is not su£fic~rrt for the purpose. 5) Iho dovelopment of analytical theories of cede trm~partat~on and deposition may be -tided by such compiled. data. For o:xamplo, If it wore find their boach sand in general Ad ~ negligible skis, it would bc possible to simplify the thooretical condi - ticks :~ecossary for the deposition of dune sad. Moreover, the absence of a skewness term In the sent would permit the direct application cuff probability theory and the use: of standout probabil ity integral tables in hauling the data. . . . . An increasing recognition Within :recent years of the variations of sediments during transport (Russell 1939) suggests that as work progresses it mar be des1:rabic for cmsidor sits in tows not Any of their environment of deposition' but as wo11 in terms of their distance =om source, if the latter call be dote~inod. Thore so~s to be little doubt that sodimontary cbaractorishcs- vary, more or less systematically with dist~co from source, Ad such canes Will be merged into the general group a~rorages of the prosont scheme of classit~cat~o;~. lIowover, once the data care assembled and available, it ma~y bo possible to red Mine the original lit Pratt in forms of transportation flaws, =~ to r~ assemble the par~etors into such secac.da^= groups. This compIexity . . . probably will not influcacc tho co~.:i ttec ' s Work in the near .tuture . General Conclusions On the ass~qptnon that the cornp~lat~ons antic~pc.ted by the co~ittec's work Viral prove of Ague to geologists, it scams desirable chit Floors co-opo~to In the ~tu:r¢ c~pilot~on of such data by taking cognizance of several features Chill greatly reduce the burden of
--- 20 ~ Gore: in~rolvod in comb long the lo tc:raturo for data. appear to bo in ordor: 33~hi~b~t r ~ J _ _ ... . . Tho fo~lo;ang sug~cstions 1) iVhenovor possible tho author s'~ould:tncludo Pueblos of his ori<gina1 analytical data, ~-rr~god in some colt class intervals based on ~ ~goomatric ratio of sizes (cithor lYontr~or.th's or Atto,rtcrg's scales or sub dim dices of them . The analyses shout d pref orably be trade with at least f~ - size grades. ^) As t~ ~ possible, open arlds in tho.~.3aalyt~1 dote should be avoided. restrictions of present techniques often Co~L+ro! this point, however, inasmuch As most Methods in cordon use do nct apply beyond ~ micron. Open finds in sieved data nay be avoided by o~t=~pg tho .~alysis ink the sub-si=o sizes. ~ practice test for open-endednoss of Clt~QS-iS that the last grade detornined has a greater percentage than the next preceding grade. The analysis should be extended z: possible until the last grade determined has less material i;han~the preceding grade. Percentage frequencies should be reported lo the nearest 0.1 percent. ~ . . , ~ . . . Then it is not feasible to include all the analytical data in the published paper, it may be possible for the author to vend ~ copy of his analytical data to the ~ub-co=~? ttee. Authors may also co-operate with the cormittee by f-Reship it with complete statistical summaries of their work, even though they may select particular r~eamres for their own public&lion. Such procedure would insure the early inclusion of the data in the su=~a~zed reports. . . . . . . . . . sufficiently detailed description of the samples or^alyzed should be included in published reports' so that as sedimentary classifications improve ~t will be possible to assign the sample to its appropriate group without ambiguity.
Whit it ~ .~1 ~ -- 21 Why Erumbein, W. C., (1936) Applicat~on-of logarithmic moments to si he frequency do stributi ons of s ed~ment s: Joker . - Sed . Petrol. ~ vol. 6, Up . -35_a=7, 1936. T`~:rumtein, ~. 0. Id Pettijohr~, :E?. i., (1938) Equal of sedimentary ,, , , _ Petrography, New York, 1938. Otto, G. H., (1939) A modified logarithmic probability graph for the interpretation of mechanical analyses of sediments: Jaw. Sed ~ Pet:co1., vol. 9, pp . 62-76, 1939. . ~ . . Rouse, H.,{1939) An analysis of sedun£~t transportation in the light of fluid turbulence: IT. S. Depth Sari culture, Sedimentation , Division, ~ 1939, p. 20. :Etussel1, P`. :D., (1939) Erects of transportation on sedimentary particles: Symposium on Roget Scdimonts, Tulsa, 1939, in. 32-47. Trask, P. D., (193-2) Orion n and onvi~~omnc~Lt of source sediments of Petroleum, Cousin, Texas, 1932. Twonhofo1, W. E., (~939) Principles of sedimentation, Ned York, 1939* iVentworth, C. K., (1929) The mechanical conpo,sit~on of sediments in graphic form: U:niv. Toga Studies in Nat. Bristol Sol. 14, ear_ ~. . ., ~ . , ~ 931.
