KETY BOOKS

---

KETY BOOKS Corrosion: Caiises and Pi.e!,e,ltion. By F r a n k *\-. ,i'peller. 23x16 cui; p p . x i i i f 6 2 1 . .\-eic I-ork n i l d Londoii: JlcGmzr- Hill Book C o m p a n y . 1926. Price: 86.00. I n the preface the aut,hor says: "Interest in t h e corrosion problem has been increasing for many years. This has inspired investigators and n-riters to such an estent t h a t it is noir difficult t o follow the voluminous technical litemtiire relating to this subject and select t h a t 11-liich is useful. One object of this work has been to take a n inventory of the avai1:ible information and classif>-i t according to a aimple and practical system to make it clearer and more readily availalile to those n-ho have not studied the ailiject in detail. I t Trill lie seen from a study of t h e various controlling factors, t h a t those n-hich act esternal t o the metal afford a natural and logical basis of classification of the various types of corrosion. Such a .system has been followed in this book. I n addition to the a~-ailableinformation from the literature, data have heen included from the experience of the author and those n i t h n-hom lie has been associated in the study of corrosion during the past twenty years." 11. viii. The book is divided into two parts : general principles and prerentive measures. Under general principles the chapters are entitled: introductory outline of the corrosion problem: nature and mechnnieni of corrosion-theories of. corro.5ion: influence of methods of manufacture antl trentmcnt: influence of factors internal to the metnl-composition: influence of ificntion: principles antl methods of corrosion testing: fnctors esternal t o the mptnl-el 1-nder preventive measures the chapters are entitled: relatiT-e corrosion of ferrous nieta prevention of corrosion in the atmosphere: prevention of corrosion under v-ater; prevention of corrosion in closed x n t e r systems by removal of dissolrecl osygen: prcvention of corrosion in high-pressure steani plants-boiler n-nter t r w t m e n t : prevention of corrosion in steam a n d hot n-ater heating systems: chemical corrosion: prevention of corrosion underground: stray-current electrolysis: classified list of selected references on the corrosion of ferrous niet 31s j appendis. "It ~ r a thought s for m m y years that n-ronght iron actually had a filirous structure and. indeed. the number of persons still holding this vien- i.5 suprisingly large. Many valuable properties were attributed to p u iron on account of its "filirous structure" which were denied to steel because of ita "c line .structure." The microscope has summarily disposed of this erroneous belief in showing that t h e ferrite constituting the bulk of u-rought iron is in no way different, from the ferrite forming the hulk of l o v carbon steel. Both are . being made by the de-La\-aud centrifugal process. The molten metal is poured s p c h r o n o u s l y from a moving trough into a rapidly revolving horizontal mold and is thus progressively fed throughout the length of the mold as it revolves. This metal has a dense fine-grained structure which must be annealed t,o give sufficient, ductilit,y. T h e texture is different from gray iron. TI-hile not so likely t o suffer from graphitic corrosion, there is some question as t o the effect this treatment \rill have on durability. This material! hon-ever. has not been used long enough to determine its st,anding compared n-ith regular cast iron. -1modification of the centrifugal process has recently been introduced using sand-lined moltis: this avoids the necessity of a n annealing operation," p. 7 1 . "-4surface n-hich has been highly polished is very much less attacked than a rough surfnce. Frequently it withstands exposure for a considerable length of time before showing :in)- corrosion, irhile unpolished surfaces under the same conditions may be badly attacked in a relatively short period. Polished surfaces usually corrode more evenly. This is particuh r l y true of high-chromium steels. The rougher, more strained, and less uniform a surface, the sooner corrosion will start and t h e more localized the action will he," p. 71. "The analysis [of the Delhi pillar] might he taken for a n ordinary sample of modern charcoal iron except that the carbon is a little high. T h e d r y hot climate of Hindustan probably has had much t o do with the durability of this particular sample of old iron. T h e

I112

S E W BOOKS

pillar ivas evidently hamnicr forged. during ii-hicli process a considerable amount of the fiisilile slag formed n-as proiinlily n-orketl into intimite contact n-ith the surface. -llt,hough positive proof has not, bec,n ohtninctl so fxr, Rosenhnin is inclined to explain t,he great durability of this pillar and simi1:ir c n s c ~IJ>tlic presence of :in e ~ c e i of s protective cinder found on the surface of the metal. S:ii~ipl~+ of such iron riistrcl frwly Tvhen (aut up 20 as t o expose a fresh surface. "In 1 9 ~ 8the , author had the, o1q)ortunit)- t o m i k e R r l o a ~examin:ition of some of t,he oltl French pipci left o n tlir Istlimii.: of 1'annni:r nl)out tn-cnty years liefore. 1Iiich of the ]mint, Sonic pi this ~ v e r ealmost us good as n e x , pparenily u n p r o t c ~ t c dIJ>s w r e almost cntirc~lytlc.*troyctl l)y rust,, :irently this vas not tluc to any ntial difference in the 1)roc ( i f Iiin1iuf:icturc~.for piercs of both wrought iron antl Fterl ivcsll :ic: in t h a t which was l,adl>-( ~ ~ r r ~ d t t l . pipe n-ere found in the m:itcrinl irliicli suivivr,tl .inalyses of man>-wniplw of T I ~ f:rilctl wt~ o intlic:itc, nnyt,liing iinusunl in the composition. It is possible that the metal ivhirh dit1 not corrotlo wns firbt oxposed duriiig thc hot dry season. a t which time a elchytlr:itetl oxide coating i w s formed proirctcd t.he steel from h t c r :ittnc.ii. I t is i n t e r c t i n g t o IiCJtC' author made on oltl piping f r o m this i o i i r c ~slio\vctl t h a t , on rc' thc old metal corrotlcc! in tlirx ntnioaplic~rc~ of tlic Istlirrii~s:it, much thcx s:inii>rate us stccl pilw niadc~by niotlern pro( "In 1923 Uiick reportid thc rviiltq of :in c~stc~niivc st,rir,s of tests on hare shect mrtnl u n t l ~ rnctunl service condition. in throe tliHcxrcnt, c.linixcters of atnioiphere. The result.: of . tl(~citlr~t1 wiwriority for opcn-hc~:irthm t l I)c this research prorctl a , l , j t o 0.3'> pt'rct.nt OVCY sirnilnr .stcxcl v-ithou iiig copper in amoiints :ind intlicatcd that tlicx Iifc. of .;iic-h iii:itc,ri:il iq niorc than tlouliletl I)y the introduction of copper. L:itrr investigntion provctl t h a t this estinintc~is vvry conscrvntivc," p. 103. "11:my invrstig:itors h n ~ cnotcd mid comnicntctl iipon the rc~m:irkaIile tliffrrc~nceliet n w n t,hrl dcnse. tlnrk-colorcd. :ind rlo~sc~1~:itlhcrcnt rust film on copp spongy. a n d light-led film 011 iio~i-c~ppi'r stc . It i.i :ipp:irent thnt thc dense rust film c ~ s nntl fairly vficicnt protector t,o the untlc~rlying ivhicah forms on coppc'r stcvl I ) c ~ ~ ~ n:Ii IWII ntl esporiinc~ntsh:rw 4ion.n that thc, c ~ ~ r r o s i or:ite n o n copper steels hec~iinc~s 1t9s? -ern1 months' t ' x l ~ ~ ~ i i r ~ , . rom t h r tl:it:i :it hand it herc is :I strong indication that the \ v ( ~ I l - k n o ~ v ~ i , influence (If sulphiir on n ji!r n!~,iospli~re] is in soni:' n x ~ n n c rc-ontrollrd 11y thc addition of ( ? O ~ ~ J C Tniid . :il?o t,lint t h r strkit'tiirc~of thcx ruqt filnl formed is a function of the, ince pcculinr t o coppcr ~ t w l . Thk diRcrc~nccin itrricturc m:iy l)cl d u e t o tlir don-er corrosion rntc. and it :ilso m-: prolnl)l(~ t h a t , :it :i 1:rtcr pcriocl. it acts mech:inic~:illy a i :i protector of tlir uiitkrlying nicstnl. It : ~ l s owems l)rol):~ldetlint the theor>-of s d p l i ~ r control, above mentioned. (loc>snot complrtcly exp1;iin the. p h c n o m ~ n o nfor , the re:ison t h a t CSY and .iston find tlint I)!ntlding coppi'r to piiw clcctrolytic iron (pres~im:ililysulphurheneficial rcxsults arc>ol,tniiietl from :I corrosion at:indpoint." p. I I C . Thcl addition of from 0.I O t o c.25 p c r c ~ n tecil)p!~rto n o r m d open-lie:irth or licsscmcr is decidetll>- 1,encfic.i:il ~ h n the, i mct:il is cxp tl without protection t o wet and d r y v-eathrr conditions. l'licx s:1111(' i- prol):ildy t , r w in :I largc nic:isiire x h e n the mc,tnl is pnintcd or c o n t r d . The maximum cffcc.t i q oI)t:iinetl :it :ilmut c. I 5 I m w n t with c.05 percent sulphur contc>nts. -4s n rule. the :idtiition of copper to stc'c.1 :ilJpc:w t u hart no tkcicied influcnce one n-:iy or thci otlicr \\-lien tlir nic)t:il is exposed untlcr ivatcr, in t soil, or in plncrs whew r. that there m'e mnny thc air doe.: not iiuvc f r w :I(YW*. I t should l i c pciintcd x-nrietiw of soil and x i t t > r ,ant1 tlint rc1:itiwly fc\v t w t s 11: 'n mudc in such mcdi:i. compnrctl n.ith th(, more ronipli~tccxpcriencr lvitli ropp t h e ntmouphcrc. I-ndcw atmospheric conditions. t h e effwt of copper is much more marked than is the purity of iron or strc,l. "11ore thorough iiiTcwtig:ition :ind study i.; rcqnirctl I J C ~ thcO vminl)!e ~ ~ infliirnce of copper in subnqucous anti ntmwphcrir corrosion run l i e explained satisfnctoril?-. T h e ad-. vantage of ropper-l)tJ:iring stccl, hoivevcr. hna h c > n satisfnctorily demonstrated unc1c.r atmospheric conditions antl in certain kintls of silt \ w t c r and a. considerable amount of this

NEW BOOKS

1143

materinl is 11on- lieing used for construction which is exposed t o such conditions. It has lieen cstim:ited that in 1923. :ipproximately joo,ooo tons of copper-lienring steel were made in thc Unitrtl 5t:ltes." p. 119. ,.Bydrochloric, sulphuric. or sulphurous acids attack stainless stecl or iron rapidly. 1-itric acsitl of :ill strength? ha. pixcticall,~-no action on such m r t d and is. therefore, a good crirerion for tlist,ingiiishing stainless from ordinnry iron or xtwl. Ckmmium steel which lias !,ern immrrscd in 1 . 2 0 spccific gravity nitric acid for 24 hour,* i d rehiat sulphuric acid rvernl houri. duc to t h c p:i.*sivifying effect of t h r nitric :irid. tee1 is nttacketl liy piirf, dilute sulphuric acid, Hatfield h in the hnrclened condition. it is not attacked in mine ivater, \yllich contains sulphuric acid nnd ferric. sul1ilinte. iinlers the ratio of wlphuric acid to ferric sulphate esceeds -steel. where it acts as a tlethe rate of corrosion. The retarding nc-tio II chromium ulloys is crensctl trn(1enc:- for t h e metal to liccomc p 'vc in the presence of ositlizing compountla." p. 126. ,'At high tenlper:itiires. chromium steels are oxidized lcss th:m :ire carbon steels, and form the climncteristic colored tempering films at. n highrr tempcrnturc. Considernlily higher tcmperatures are rcquircd t o g i n >the same thickness of oxide on the stainless steels n-hich do not nctunlly liegin t o scnle I ~ ~ , l o~n -g c o t"o 1 8 ~ i o ~(X16' F to c)X?'C'l depending upon the amount of chromium present ," p. 13I . ,'Turlinlent floiv in pipe lines ant1 agitation of solutions in open T iisiiall>- I r i n g nliout a riinrkctl increase in the rate of rorrosion. T h i s i thc amount of dissolved osygen reaching the metal surfncc i:, usiidly a maximum under t h m e conditions. I,( thr r i h t TIcc ( ~ , f o . ~ ( / q cif/ ! the , motion of the solution in contact with t h e snrfacr is not uniform. thc area sulijcct to the higher ve1ocit.v liecomes anodic with local solution of the mctal. U'hcn dissolved oxygen i- present. t h e polarity d l usually ~ J Ct h e rwerse as the nrcn Ivhich id in contact iyith solution :It the higher velocities received a greatrr nmounr of osygen antl therefore tends t o liecome cathodic.," 11. 149. ' $ T h e humidity of air nlmve the corroding solution influencc,s the rate of solution of gaseous osygcn 1))- quiet miter. The rate is grc:iter for dry t h a n for humid air. This is d u e t o the incrensc. in the rat? of eraporntion from thcl liquid surfarc, as the :iir :il)ove it Iiwonirs dryer. Ev:iporntion caools the. siirf:ice layer of liquid so that it t)ecomc.s more densc a n d sinks, thiis clistri1)nting the disqolvetl ox,~-gcn( n h i c h wi:, :idsorlx~lnt the surfncc, I through olution, antl Iiringinp fresh unsaturated liquid to the snrfnce. This phenomenon causes even more rapid ox,vgen nhsorption i n d t solutions xvhere, Iiecaiisc. of the evnporation taking p l n c ~ t,h e surface of the solution is concentrated. Thus, in :ititlition to the cooling effect. the concentration effcct still furt,her increases the density of the liquid a t the surfnce ant1 aitis the formation of ronvcction currents. This effect re:irhcs a maximum at a coricrntrntiori of Liliont one percent in sodium rhloride solutions," 1., 192. ,,High tempernture ositlntion appenra to be due mainly to the tlirect chemical comhination of oxygen nit11 the metal; whereas the artion may aln-o>-slie considered to he electroclmnical n t lower tcmperaturrs in the presence of moisture." p. 193. Pillirig and Beckworth "divide the metals into tn-o groups:fn:those which form oxider linviiig a cellular or porous non-protective structure: m t l 1 bl others which form a continuous protective oxide. The former include most of the light metals n-ith low-melting points. such as mngnesium ant1 canlciuni. while t h e latter include copper, nickel, chromium. nluminum, and iron, They h i v e shown experimentally. especially in t h r case of copper, nirkel, and some of the coppernickel a l l ~ ~ ytqh. t the rate of oxidation n t any ternper:lture is tiependent upon t h e rate at Ivliich oxygen is trnnsferred through the oside layer. In the case of copper, this is thought t(3 l i e tiependent upon the rate of solution antl diffusion of osygen through this Inyer. The r a t e of ositiation decreases v-ith the time of esposiire [ a t a constant temperature) in the form of a simple quadratic parahola. thr amount of oxidation being proportional to t h e square root of the time. This is nnalogocis to the retarding influence of natural protective coatings deposited from water which may follon- a similar law," p. 194.

1 I44

NEW BOOKS

“The system of classification n-hich has lieen developed and used throughout this !>oak includes five separate classes: I 1 atrno.sphcric corrosion; ( 2 1 underwater corrosion; i s i underground corrosion; (11 chemical corrosion; i j rlcctrolysis (due t o a superimposed current). Each is distinguished l)v the controlling influence of different, nell-recognized. external factors. Thc outstmdiiig chnracteristics of carh class or type of corrosion will 1)e taken u p briefly. “Atmospheric corrosion is c~harnctcrized13)- the presence of osygen in excesz, moisture being present only part of the time. ‘The metal is ivettcd and t l r i d and is subjected t o t h c action of sunlight and varia1)k Tw:itlier. AIetal roofing. fence wire, l)ridgcs, rails, str,cl tanks, and cars, are examples of matcrinl ~ i s t dunder such conditions. -It lenst eight!cent of the stcel in use is sulijcct to this class of corrosion.” 1). 195. ‘T7nderwater corrosion in 1ir:ictice inrlucles m t r r conduits of all kinds, lieat steam Iioilers, submerged iron and stccl n-orks, nnd other surh niutcwials. applies to ten percent of tlica f c r r o u mcltnls in use. l’hc dissolvrtl oxygcn usunl for corrosion t o continue is oi)t:iincd origiiinlly froin thc, nir m d is limited in iini the conditions are the reverse of tliosc in t h e ntmosphcrc~ivith respect t o the ratio I)ctwwii the nvailable supp!y of oxygcn ant1 n-atcr. T h a t is to say, in t h r ntmosphcrc, polnrizntion by hydrogen is preventcd l)y the lnrge c s c w s of oxygen, ~ ~ l i e r e ni nb witclr the iiictal surfnce is a l n a y s more or less poliirizrd. “ I t is not surprising, tlicrcforc,, t o find th:it untlcrn-ntcr csorrosion diffcrh n-idel\- in n ~ t i r l y every respect froin atmospheric3 corrosion. T h r prodtwts of rorrorion usu:ill>- contain aliout one-third ferrous liydroxidc, :is tlir nictal is incomplctely oxiclizeil due t o the lack of oxygen, a n d tlic rust is soft and more uniform in trxturi,. T.-ncter 4niil:ir conditions. ncithrr tht. amount nor the structurc of the rust formed is influcxncwl t o nny apprwiaI)lc d ~ g r r1-, ( ~ tlic, ordinary changes in com1)osition found iri commrrcinl iron : ~ n dsteel,” 1). 196. “Tnderground corrosion, wlicrc c+x*rrolysis froin strn!- currents doe3 riot cntc’r into thcl problem, constitutes :I class which com1)incs some of the ch3r:icteriatic.; of c,ach of t h c tn.o foregoing classes, but whiclh also exhil)its ccJrtnin charnctrristic3 peculiar to itself. Something of the importnncc, of this clti.5~of corrosion ma!- I)(, judged from the fact thnt morL’ than half of the pipe manufactiirctl in thi.; roiiitr>-. :iinouiiting t o over :I iiiillioii tons annually, is used underground in the conPtruction of n-:iter. oil, nntl gas liiicls. Besidrs pip:. and other structurcs underground. this class mny inrludr thc rorroaioii of frrrous metals in concrete, in cinders, or in nny caw n-hrrr the iiictnl is c~iic~lo.wd I)>- a more or less solid Inass cap.icI)le of containing m t c r , ” 11. 197. ”Considering tlic relntive influencc of fnctors intcrnal a n d external t o t h e metal, it seems clear t h a t , with fen- exreptions, cJ.;tcmi:il f’ictors predominate :ind determine the ratc of naturnl corrosioii with ordinary iron and steel. C’omposition is of minor importance in corrosion underwater and, with the notal)le cxception of the, copper contents. the composition of ordinary rolled iron and stccl is also relatively unimportant in thP air. “The mass of evidcnce whic7h has been nccuniulatcd on tlic rclativc corrosion of wrought iron a n d steel, clearly points t o this generd ronc.lusion which might have beeii predicted from the present knowledge of The influence of t h e w factors. I t is also true that the r e l n f i m corrosion of metals varies greatly undw diffrrcnt conditions. This is strikingly illustrated by the f a c t t h a t in the air,, copper-bearing steels corrode much slower than ordinary iron or steel, hut in fresh water: there ia no appreciable differcncc in the corrosion rate of the two. Generalizations should be avoided in considering rclative rorrosion,“ 11. 267. “;lmerican vermilion [basic chromatc of lead) gave better results in the Atlantic City tests t h a n a n y other single pigment te,i.trtl. having a rating of practically roo percent even after four years of exposure. Instances have, moreover, been cited where this pigment used upon iron surfaces has giren protection from rust for periods greatly exceeding ten years. I n subsequent tests a t Atlantic City arid in Kashington. this pigment gave very remarkable results, showing greater resistance to wear than any other paint tested. Unlike red lead, it does not neather to a whitish srirf:ice, heing apparently unaffected by the carbon dioxide in the atmosphere. Due, h o x c r e r , t o its relatively high cost as compared to red lead, its use has been restricted as the sole constituent of metallic primcrs. Severtheless,

NEW BOOKS

1 I45

it is being used very satisfactorily in relatively small amounts, say from ten t o fifteen percent, in many metal primers. A wider use for this very valuable pigment should follolr-. *hiattempt was made a few years ago b y -1.S. Cushman and H. -1.Gardner t o surface-coat the particles of certain pigments ti? a chromatizing process. By this process, there vas formed around the pigment particles a coating of basic chromate of lead. Kliile the pigment thus formed is not as durahle as 100 percent chromate of lead, it has, in various expcriments. giren excellent protection to metal over a long period,” 11. 292. “Zinc oxide is a strongly rust-inhibitive pigment, due to its basic nature. This influence asserts itself ivhen used in combination with lead pigments in light-tinted paints which are to some extent used for finishing coats. One of the tiest metal paints for general u.jc can lie made with a mixture of iron oside and zinc oside, the zinc oxide greiitly increasing the rustpreventing properties of the paint. This t>-peof pxint is finding quite n wide application in railroad n.ork and is giving gencr:ill?- good serriee. Zinc dust. liecause of its position in t h r electrochemical series, is a n excellent protectire of iron and steel surfaces, t o som” (,ut cnt acting likp galvanizing. I n fact, a so-called “liquid galranizing” has lieen prep:ired from :L finely divided zinc dust composition in boiled linseed oil. Paints made n-ith equal amounts liy n-eight of zinc dust and zinc oside are remarkalil>- efficient as primers for met31 surfaces.” p. 293. *‘.innluniinum paint film has a n u m t m of unique properties. I t s bright metallic surfwe is highly reflecting. The metallic aluminum finks are entirely opaque and hence protect t h e underli-ing vehicle from the sunlight, which is n most important factor in causing deterioration of paints. .Iluniiiium powder has lieen shown t o considernlily increase t h r water-resisting qualities of the paint film. Before applying aluniinum pnint t o steel, it i j desirable to paint the met:il n-ith ‘i priming coat of :I good Ixi.;ic pigment. utilizing the protectire proprrty of the aluminum in the finishing m a t s , ” p. 29j. “The phosphate coatings are somen-hat simi1:ir to oxide coatings. Thcy are producctl liy the action of several combinations of chemicals, a11 of n.hich produce iron phosphates on the surface of the metal. I n general. tlicse coatings are slightly lcss cathodic t o iron than the oxide coatings, and, therefore, are more protective. This fact, caoupletl n-ith thcir better adherence t o the iron surface and thclir pleasing appearance, has given them :i rnthcr extensire application Like oside coatings. they are made more protective 11y trentrnent n-ith a suitallle oil. -1sin other coatings of this nature, holyever, the>- are not sufficiently resistant t o lie used under exposed conditions,” p. 3 0 2 . “Steel embedded in concrete exposed t o salt air ncnr the sea is more auhject t o attack than in concrete completely submerged in the sea n-ater. This is clue to the fact that moist air d l penetrate through porous concrete t o the niptal, while salt w t e r niny not. Corrosion of t h e reinforcing metal, corered with tn-o t o four inches of good concrete, has occurrccl above the high-water level of concrete piers in t h e sea. Tlie volume of the rust formed is ahout three times that of the metal, so that its expansive power is usuall?- sufficient to c a u w spalling of the concrete. I n the same locations, no trouble has been experienccd with t h c concrete under the sea n-ater. I-xidrr most atmospheric conditions, however, t n o to four inches of ciell-mi;ced concrete is sufficient to give permanent protection. K h e n not directly exposed t o the neather, as in the protection of structural menitiers in buildings, a corering two inches thick will usually suffice. Where reinforcement is necessary, in concrete exposed t o sea air. the metal used for this purpose should lie painted with ttvo thin coats of a red oxide of iron paint made up with China-wood oil nntl spar varnish, or dipped in a bituminous coating, and then dusted with sand. .liter all the best protection will be ohtained b y using a dense concrete of proper proportions, thoroughly mixed and well tamped around the liars t o a sufficient thickness,” p. 304. “d petroleum grease misture carrying about 65 percent b y xeight of zinc dust) fine enough t o pass a zoo-mesh sieve, has been dereloped in t h e Research Laboratory of the S a t i o n a l Tube Company and is used t o protect pipe threads against galling and corrosion. A little sodium chromate is desirable unless all n-ater is eliminated from the grease,” p. 307. T h e process of calorizing was developed “as the result of reseerch in the General Electric Company Laboratories in 1911, in a n attempt t o protect iron from oxidation a t high tem-

I 146

NEW BOOKS

pcmtures when the metnl \viis hcutctl under oxidizing conditions. .iluniinum oxide is formed on the calorized surface :ml. to :a largr extent, protccts the metal from further oxidation as long as the temperature docs not c ~ c r e c l17oo'F (927'C). At this t,emperature the protectirc film of oxide begin.; t o I)reak down. Figure 66 indicates the rate of oxidation on pieces of cnlorizetl s t w l tuliing compnrctl with untrr,itcd tuliing a t 18\30 and 1475 'F (989 tinti 802'C I . Calorizing h n s h e n found useful in provriiting th(, deterioration of superheater tulles, pyrometer tuliw. nntl cc'rtniii oil rcfinery quipnic~iitexposed t o high temper:itures and corrosive sulphur cornlmiintl- siirh untl in prcxisure >tills," p. 320. "Large steel n-iiter mains h n r c bccn . illy wnicnt linrd nithout reinforcement 11)placing n concrete n.ootlen nioltl in the' poming the cement mixtiire through temporary holes macle in the top of tlic, pipe. Steel srctioiis of thc Xew T o r k C'atskill Aqueduct \\-ere linrd in this way n-it11 n I to 2 siantl-port1:intl cenicant mixture tn-o inchcs in thickness, I t is most important in all c H, of m i r s c . t o protcct tlw ccmicnt from loss of n-atcr by evaporation until it is thoroughly r r t . 'I'hcsc, large pipes w i r e protected and rpinforcetl on the outside liy :I jacket of 1iortl:ind cclnient concrctc'. J17hcxre organic matter dit1 not form on the interior surfacc of tliv pilw, t h e c.cirnentIativcl\- little retliiction in flow c:ip:icit,y ~ h c tcstcd ~ n :iftcr five y ~ m onditions, the, iiisitie n ordinary asphnltdiamc~tcrof ccnicnt-lincd pi1w mny l w figur(~1 m a t e d pipe to secure thc s m i e rat(' of flow ivithout cl(ciii.qing ovor :I long p:'riotl," 1). 3.16, tihfartory protcwion in w r y pure>w i i t m . ycat ".4ltliough sodium silics:itr usually gil it has lwen found that thv prcwiice of i~:ilcirini: m t l inagnc~siumi:tlts. pxrtica1arl~-the r:irhonntes, greatly aids in the, ;ormatioii of sii(~1ico:itinqs. JJ-hilc mort or less porous in chnractc~r.t1ic.y :iw iis~i:all~rufhcic~iitly tlistriliiitcd to retain t i high frrroii--hytlrositl(~c,oiic~ciitr,itioiin ft,re with the diffusion of osygc'ti to thc nwtiil surface," 11. 3-c~). ' . q .ern1 ycnrs cxperieiiccs in thc. tlcsign anti in?tnllation of nearl\- 2 0 3 pl 1-nited States for the r e r n o r d of oxygen from witer hns tlcnion.;tratrtl of appirntus cannot lie uscd for :ill ptirliosw. The choicc of the Ixst i.;ting set of conditions depends, among other thing?, upon the mnsimiim hourly rate of flow of wiiter, the numlicr of high 1 x ~ i klontls :it n-1iic.h \\-;iter is rcqriirid for use. thp st!mn > ~ p p l > iind pressure avnilalilc, how inwh resitlunl osygPn is pcrmis*il,le. :and whclther or not a n engineer is aln-ays on duty. For thc hot-nntcr supply of 1)uildings of nioderxte size. using up to aljoiit 1,000 gallons per hour. the tlc:ic.tiv:rtor, with or 11-ithout n filter, is usun!l!- the most convenient nnd economic.nl npi)liancac to IW. In larger liuiltlin power plants, a mechanical c1cnc~r:atc.rof sriitalik tlesign \vi11 take u p le more economical and efficient. "Khere witcr c>onditioiis:ire such that the protection of both the cold :ind hot !vvatc,r piping of n building is necesbary. it m i y Lie most cwmomii-nl t o reduce the frce oxygen in d l the n-ntcxr entrring the liuilding to one cwl)ic centimeter p p r liter or less by a vacuum deaerator operating a t normal tempxxture a n d , after hextirig t h l water for the hot-water supplj- system. t o remove the I,:ilancc n-ith :I c h c t i v u t o r or anothcr ni:~chanicaldenerstor. By properly proportioning tlic plant, it should Iw possilile to regulate the residu31 oxygen in the cold and hot ivntrr YO tlitit the life of thcw sJ-atems n-ould IIC alJout e q u d :md would lie commcnsiiratc Tvitli tho o x p c t i d life of the Iiuilding," p. 396. "In comp:irntively sinal1 liuildings. where thc. lcngth of the, piping system is not over 100 or zoo feet or where vcr!- large qunntitics of wLitwarc used, oxygen rrmovnl :ipparatus may he too costly to install. In such c a s ~ s . consicbrnlilc , protwtion and clear rust-free sodium siliratc," p. 397. ften lie olitaincd I)\- the addition of limr t r m s conlti unt1oul)trclly lie realized vings in rctlucctl tlrprccintion of hcnting and expensive rcyairs avoided, if 1)uiltling o\vners ~vouldoccasionnlly ernploJ- nn industrial n-ntcr chemist t o test t h e quality of the n-atc,r nntl s t m m and to advise o n the conditions of the system a n d the u--att>rtrcitment required. Ailimited nmount of such servicc s h o d d cost approximately little. Such tests involve the dctcrminntion of free oxygen, carbon dioxide, a n d pH 'i.nlue of thc, steam entering the system aiitl of the condensate from the returns, in addition t o the usual anal\-.ses of the feed water and the ivntcr in the Iioilers," p. 456.

S E T BOOKS

114;

tems occurs inside the short picces of pipc "l\Io.;t of the corrosion in steam-henti ~ n i p p l r slentling ~ from the rndintors. The ild prcfer:il)ly lie rnntl(, of cstra stronq pipc. In siic~hsystrms threadctl joints are niinieroua nntl may permit the infiltmtion of air, so t h a t prcjt carf' slioulcl lie ~ n k c i ito sec that all joint.; nntl vnlvcs :ire mntlc, ant1 maint:iined air tight. Pipc threads shoiiltl I r clcnn-cut with good dic-5 7 0 :is t o makc R tight joint. Too little att-ntion is pnitl to this m a t t i ~ r . "=\ siiitnble threntl 1uI)ricnnt i,: :I n ity. as the, ortlinsry thrcatling tli. tloci not cut i i threntl n-hich is gns tight iinlc~~s jo111c of tlirc,atl filler i,+ u d when thtj pip(>is instnllrtl. The orclin,iry n-ell-knoivn red leatl-glyc~erinc~ niistiirL>is good, 1)ut it h:?' Iwen foiind thnt nny of tlie tlircntl luliricnntj can he) ninclc t o Iirotlucc n tighter nnd more t1ur:iIilc joint l)y the

illric~antfor w e o n

the!-

scll. Th(L time i p 1)rol)nl1lycoming \\-lie~n.tc:im will I)(. sold o n :I *pc'eifir:ition ~irc,~ t w l mnl. . a n d o t h w c~onimotliti~w.On the other liantl. it woiiicl s c ~ i i i ntl orc*liitrct.;~hoiildp:iy mor(>nttcntion to the c~onstrui*tion of dtc:ini it :ill connections n-ill Iir m:itlc~ :is tight :L< 1msilile, ivitli suffieiont grxtlc t o ( w r y off contlcnuntc~.a n d iyitli 1lropcr iirotection of t h c c,stcrior of the pipinn . .. d o n , 1). 462. ,'Cast iron is c.orrodtd very little I)? fiiniing .;i111)liuric :witls; hut n niinilicr of c:i-r< hnvc i l c c ~ n r t ~ o r t l r t lw1ii.r.c. tlic m(,t:il has crackrtl i n pr:irtiw rvith nlrnost ivqilosiv(' violenec~. While this ivas fornic'rly .lttril,ntctl to tlic prwsiire pr(ot1iicrtl I)? .siilphiir dioxitl~:ind hvtlrog-cn fornicti liy c~orrosionIictu.ccii tli? crystalline prnins of t h r nicat:il, niorc rcw~litv o r k hy I3nni:an. ha.' intlicntcd t h a t frcr sulphur trioxitlc, rcncts rcx:idil\- ivith silicon i n the, mc,tnl. forming silicon dioxide, n-1iic.h owing to t Hnmlin nntl Turncr state that in ortlcr to nntl Ion- grnl)hitic cnrlion content." p. 479. "The ni:ijor portion of corrosion nntlergronntl takes place on iron nntl steel tul)iilnr lxoclucts, so that much of the tlisciission of soil rorroyion n-ill ncct~swrilv1x1 1)nsed upon record5 and esprriencrs with 2nd oil lines. On the whole. espc~ricnc.ciiitlicntes t h a t the life of iron nntl strcl 1 lergrountl. h:w Iwen c~onsit1eral)lyundercs~imntctl. Tho nctniil rcplaccnient rrrortls 1 iron n n t l h t w 1 oil lincs more t h a n a hmiclrctl milcs long, tr:ivcrsing irliat might l w s unuru:illy corro-ive soil. show :in averazc nnnu:il rtylnceinent of only 0.43 of I p~ he line o v c ~a period of 2 5 to ,io yrnrs. Scnr1)- all of thwe rep1accment.s occurre er linlf of this period. 1-nder more fnvor:il)le condition. in ninny pnrts of the arc knon-n whclrr hundretls of miles of *tee1 pipe iisetl for tr:insporting oil :intl gns slion- prnc~ticnllyno corrosion i n that lrngth of ~ iron nntl atcc,l pipe 1inc.s hnve IICYII time. :ind some instnncrs are on record n - 1 1 ~\vrought i I i use for forty ycii-6 with practienlly no action on tlie metal tliic to t h e ?oil." 11. 495, "In some kinds of soil. cnst iron is s ect to n peculinr form of disintegrntion knon.n as "prnphitization" or "graphitic softenin This action ih the result of t h e wlcctive corro>ion of the frrritr IIiure iron grains) of the mrtnl. Thc ~ke1etoiistructrlrc n-hich remain. ccmists mainly of thr products of corrosion nnd the non-corrotlil~lematter in thc (Inst iron. i.e.. tlie graphite, rarlidc. silicitlc, a n d osidcs of iron, and xhnt remains of thcx nncorroded met:illic iron. This structure retains the form of the original metal, hiit is soft enough t o 111, easily cut with n knifc. If left undisturllctl. it is often strong enoiigh to carry \\-:iter or gas under Ion- pressure. hilt is 1 i k ~ l to ~ - fail under shock. StraJ- electric currents sometime; which

I

1148

K E W BOOKS

produce this effect, but it seems t o occur more commonly in certain corrosive soils where electric currents are absent, particularly in those soils which carry calcium sulphate or similar salts. Graphitization also occurs in sea n-ater," p. jog. "The rapid deterioration of the condnit carrying hJ-drogen sulphide water has heen exthe layer of elementary sulphur a n d plained as the result of the contact effect set up iron sulphide which collected on the surface of t h r mctal, since these suhstances are strongly cathodic to iron. Lalioratory experiments. in this investigation, indicated t h a t the rate of corrosion vas increased as much as joo prrccxnt ~ r l i e ncast iron x i s placed in contact with the deposit from the inside of the pipc iiron sulphide and elementary sulphur! and immcrsed in the Nondorf n-ater. The mcc1i:inisni of thc action has nlso heen esplained as due to the fact thnt the hydrogen sulphide preeipitntes iron sdphicle, rind in this way dcereascs the the corrosion. The increascd acitlity of metal-ion conccntr:ition sufficicntly to accc~lcr,iti~ the unaernted solution, clue to thc, hydrogrn siilpliidc and c~nrl)oncliositle rontents. may nlso have been :i factor of importance. TYlintc3ver the niechnnism of the action. the iinportant point is thnt it i r n s m u c h niorc ra1)icI in thi, presence of dissolved hydrogen sulphidc, presuni:ibl>-in tlic nliscncc~of osypcn. This ni:i>~l i n ~ I Iic:irinp upon the severe corrosion sometimes found in oil n.ells n.1iic.h i i u d l y c:irry sal n-ntcr often with more or less dissolrrd hydrogen aul~iliitlc:ind corlion tliositk," 11. 510. These cxtracts s h o n lion- good the I ~ o o kis, Thvrc arc' some points-not m:iny--n-hic.h do not seem u p to stnndtutl. Ferrous ositlr or hytlrositlc~in alkaline solution is oni' of thc zing agc'nts 11-e liar-r a ~ i dit is tlierdorc improlxi1)le t h a t a film of thii nplircci:il)lr p i r t in p r ( v n t i n g t h e corrosion of iron in a l h l i , lip. 39, :i pity to :iwrilicl t o lI-:ilkrr, 1). 19, thtl ntldition of acid t o the r:ithodv cr to i'iiusc corrosion of th!, :uintlc, whcn Ost\vald hnd c1c:iretl 1111 t h i n point more thnn fifteen yc:irs cnr1ic.r. Tlic :iuthor cvidcntl?- has not rcntl any of the Iicttcir n-ork on passivity, pp. 190. 354.:ind 1ii.q v i m s on the sut)jc,ct arc quite primitive. T h e revimv-cxr d o ~ snot :igrec iritli thc st:itilmrnt, 11. 2 7 5 , thnt "if the electroc~liemical theory is the corrert esplnnation for corroiion in i n t e r , it naturally follows thnt it is also true for corrosion in t h c :iir. T h r only tliffcrrni~c~ i.i thc relatire importance of the various controlling factors." ?'his does not follou- :it nll :ind it mikes one ironder i\-hcthcr the elrctrochcmicnl theory is reall>-much mort thnn :I form o f words to the author. One cannot criticize the hook on the gruuntl t h a t it is merely a compilntion of tlnt:i. I)(,cause the author :rclniits quite frankly that that irw :ill hc could do; lmt the fact that it is apgnrently quite impossil)le to do much more th:in rhronicle tc,sts is damning commentary cn the n.ay in n-hieh the tcchnical people have n-:istcd their time in thc atudy of corrosion. 1T-dder D. Bnricrojt Recent Advances in Physical and Inorganic Chemistry. Bi/ dlf'retl I T * . Sterrnrt. F f j f h edition. 14 x22 mi; p p . 312. Loridoii: Lgi/giii- lines in somc cases,, they are scarcely parallel enough t o make a satisfactory grating. Pfund has shown t h a t only a portion of the color of mother-of-pearl is due to a grating structure. On p. 594 is the statement t h a t the scattering of light by turbid media is a diffraction phenomenon; hut it is not shown t h a t this view is consistent with the fact t h a t the short n-ave-lrngths are the onts t h a t are scatttred chiefly. Thc reviewer n-as intrrested in a paragraph on p. 614. "Fluoresrence and phosphorescence intrrcst tis also because they make the suhstanecs sources of light (luminescent paint for instance.) Such luminesccnt sources differ from the ordinary sources of lizht in that t h e temperature is relatively low; they emit light almost without heat. Other luminescence phenomena are, for instance, the light of the firefly. the electrical discharges in dilute gases (Gcisslrr tuhes), the light obtained on heating certain crystals (fluorspar),etc. T h e \vaves sent out hy thrse substances are almost exclusively light-waves in t h r popular sense; these sources of light are therefore very tfficicnt." T h e author does not include salt flames in the Bunsen burner under chemiluminescence, as he really should. On p . G 1 9 one would have liked a little more definite statement as t o the emission of light by a colored solution at ordinary temperatures. I t scems to be that they emit n o risible light hecausr a hlack hody pmits no apprrciahle light at ordinary temperature; but it is not stated clefinitel>-or clearly. Since a I h c k body does emit infra-red rays a t ordinary trmperaturrs, it stems as though it should be possible to photograph the spectrum of a potassium chromate solution in the dark if one gave a sufficiently long exposure. If not, one would like l o know whether the troulile is in tht. photographic plate or in the theory. ITilder D. Bancroft KIistalle und Rb'ntgenstrahlen. P . P. Eualrl. 24 X 16 cnz; p p . i2$327. Berlin; Jirlizts S p r i n g e r , 1923. Price: 86.35. This hook is liasrd on a course of lectures given a t Munich a number of yrars ago. The chapters are entitled: atoniic theory; fundamental crystallographic conceptions; crystallographic structure theory; intrrference; Rontgen rays;

XEW BOOKS

1151

outline of experimental methods; the Bragg process-spectroscopy; interference in lattices and the base nunibers-determination of structure from the Bragg photographs; Laue’s method and the interpretation of the Laue photographs; the formation of the Laue images a n d the control of structure; the Debye-Scherrer process; complete diagrams, structure of fibres, constitution of metals; representation of the structures; geometry of the lattices; ion lattices, isomorphism, solid solutions; chemical bearing of the interpretation of t h e crystal structures; lattice forces and the properties of substances. The mean distance apart of molecules or atoms in a crystal is only a few tcnths of a micromillimeter, p. 34. The mean wave-length of visible light is therefore a t least a thousand times greater than the distance lietween the atoms in a solid. From the description in the t e s t , p. 99, it seems difficult to believe that molecules of calcium fluoride do not exist in the crystal. I t is very surprising t h a t standing for eighteen days a t ordinary temperature should change the crystalline structure of rolled tin as much as it appears to, p. 144. On p. 186 there is a paragraph as t o how one can tell a n ion lattice from a n atom lattice where atom is used t o signify the absence of a n electrical charge. According t o Ewald one has a n ion lattice if thP crystal shows residual rays, pyroelectricity, or piezoelectricity. T n der certain favorable circumstances one can also demonstrate the presence of ions from the X-ray photographs alone-hecause the amplitutde of the n-aves going from the single atoms is proportional to the numher of electrons.

TVilder D. B a n n o f f Mikroanalyse nach der Mikro-Dennstedt-Methode. B!j Cnsimir F u n k . 22 X f 6 cm; Mu71ich: J . F . Bergmarcn, 1925. T h a t the method of the llicroanalysis of Organic Compounds still awaits the advent of its Liehig is a fact appreciated b y the many who have failed t o achieve success in their attempts to follow the methods of Dubsky, Pregl, and other pioneers in this field. The d u t h o r of this pamphlet had the privilege of seeing the process carried out in Pregl’s Ialioratory; hut. after a series of failures extending over a year, as led t o abandon this method of effecting the combustion in favor of that of Dennstedt, a n d in some fifteen pages he gives an admirably clear description of his llicro-Dennstedt method which he claims t o he superior to the Alacro-method. Time only will shon- if other, and less experienced. workers can use this method with equal success, or whether the year’s apprenticeship in the older methods is a necessary prelude. M. A . W h i t e l e y L’HydrogSne et les Gaz nobles. By J . J . T’nn Laar. 23 X 25 cm; pp. 7’9. Leidertc A . TT. s i j t h o f f . Price: 3floriu.s. T h e publishers of the Recueil des T r a v a u s chimiques des P a p B a s are planning t o issue a series of Theoretical and Practical llonographs in French or English four times a year. T h e monographs will he about eighty pages in length and will include important researches by well-knovn authors. The first in the series is this one b y van Laar. This is a compilation of data. including atomic weight, molecular volume, molecular and atomic weights, boiling-point and triple point, critical data, vapor pressures, densities of co-existent phases, dissociation, molecular attraction, etc. W i l d e r D. Bancroft Fortschritte der Kolloidchemie. B y Herbert F r e u n d l i c h . 22 X 14 o n ; pp. 109. Dresden and Leiprig: Theodor Sfeinkopff. 1926. Price: 5.60 m a r k s . Freundlich considers t,hat the time has come when we must make more definite our conception of adsorption, electrical potentials a t surfaces. coagulation, state and shape of colliodal particles, etc. K i t h this in view. t h e eight chapters are entitled: adsorption; electrokinetic potential; adsorption. valence, and coagulation: rate of coagulation; the stability of hydrophile sols; the state a n d shape of colloid particles; the absolute value and changes of the surface forces in colloids; phot,odichroism and related phenomena. Trauhe pointed out t.hat the lowering of the surface tension of water b y organic substances is greater t h e longer the carbon chain. Freundlich claims t h a t this can only be true

1152

NEW BOOKS

in case the organic chain lies flat in the surface of the liquid. I t seems t o t h e reviewer that, one could get a similar result. at least qualitatively, if t h e molecules st,ood vertically a n d t h e longer ones had t o be sucked in farther. To explain the adsorption from these solutions by charcoal, Freundlich postulates t h a t the affinity between water and the organic compound is the important factor and not any special affinit.y between the organic compound and charcoal. I n view of the very specific adsorption of gases b y charcoal, this conclusion is not convincing. There is a very interesting account of Karburg's synthesis of blood charcoal. pp. 4-8. The addition of a silicate t o sugar charcoal gives a material which adsorbs amino acids without accelerating their oxidation. The addition of iron causes increased oxidation of the amino acids and this reaction is poisoned by prussic acid. Freundlich defines a s the electrokinetic potential the potential difference which plays a part in cataphoresis and electrical cndosmose experiments. This is not t o be confused with the thermodynamic potential n-ith which Sernst's theory deals. Freundlich considers the electrokinetic potential as dnc t o an ac1sorl)ecl film of finite thickness and varying concentration. The properties of this film may vary considerably ITith the nature and concentration of electrolytcs in thc mass of the soliltion. Freundlich breaks a lance in favor of the view t,hat a trivalent ion is not necessarily adsorbed more strongly than a univalent ion even though it causes coagulation a t lower concentrations. This is a pleasing half-truth. I t may be true if the concentrations are expressed in molecular n-eights; liut it cannot lie t m e if the concentrations are expressed in equivalent weights. .ictually the variations are iwiially far outside these limits. Freundlich still c-lings t o the belief t h a t the amounts of the coagulating ions ivhich are carried down are equivalent. This is not true experimc~ntally.:is Keiser has shon-n, and is only true t.heoretically in case the coagulum has no adsorhing action of its OTVII. Freundlich considers t h a t the stability of hydrophile sols ir due chiefly t o hydration and he ascrihed this explanation to Kriiyt. H e docs not explain what he means by hydration. If he means the formation of definite chemical compounds, he is unquestionably Tvrong. If he means water peptization. there is nothing especiall!- new about i t . -4t t h a t he does not give 11s anything very definite as t o why tannin and gelatin? precipitate each other. It is very doubtful also whether hydration is the important favtor in determining the viscosity agglomeration t o chains or threads is the essential thing. of hydrophile sols. Freundlich ascribrs t o H n l m , p. 67, the statement t h a t precipitates are more nearly amorphous, the more rapidly they are formed. 1-on 1T7eimnrndid not say exactly t h a t betalline; but he certainly showed t,hat one gets a cause he believes werything t o bc ( gelatinous m a s v-ith very rapid precipitation. IT-eigert, has shown that in the production of photohalides, there is not only n color iidaptation but a polarization adaptation, p. 95. Glass plates were covered with a thin film of silver chloride in gelatine. This vas dried and e s p x e d t o light until it was bluish red. When this photocliloridc layer wis exposed t o polarized red light, there \vas formed a red spot which was dichroic and doubly refracting. One cannot get enthusiastic over this hook. Theory has never been Freundlich's strong point and he has ignored consistent,ly and apparently deliberately all non-European work. Colloid chemistry is not noiv, and is not likely t o lie$a German science. W i l d e r D. Bancrojt