36 



at fjerne det vedhængende Søsalt. Ganske fuldstændig sker 

 dette først ved Kogning, men ved Analyserne er forresten 

 en Behandling med koldt Vand tilstrækkelig, da det kun 

 er forsvindende Mængder af Salt, som bliver tilbage i Leret. 



Bundpro verne tørredes derpaa ved omtrent 100° C. 1 

 Nogle Gram udtoges og glødedes i en Platindigel i Perrofs 

 Ovn til konstant Vægt, for at bestemme Vand og organiske 

 Stoffe (Glødningstab). Af den samme Prøve afveiedes 5 

 eller 10 Gr. — alt efter Lerets Ensartethed — og kogtes 

 i en Kolbe med 80 eller 40 Kc. fortyndet Saltsyre (20 

 pCt.) i omtrent et Kvarter. 



Ved nøiagtig at anvende denne samme Behandlings- 

 maade ved de forskjellige Analyser faar man nogenlunde 

 overensstemmende Resultater for den dekomponerbare Del 

 af Bundprøven, især ved de finere Slamarter (Biloculinler). 

 Ved nogle Kontrolanalyser, jeg bar udført for at forsikre 

 mig om Rigtigheden beraf. bar jeg ikke faaet nogen større 

 Variation end 2 — 3 pCt. 



Filtratet fra det saltsure Udtræk fortyndedes til 500 

 eller 250 Kc. Med en Pipette udtoges af Isløsningen 50 

 eller 25 Kc. hvori Jern, Lerjord, Kalk og Magnesia be- 

 stemtes ved de sædvanlige Fældningsmidler. 



En foreganende Inddampning for at fjerne den oploste 

 Kiselsyre blev tildels udført, men er forøvrigt unødvendig, 

 da denne •Bestanddel af Bundprøven kun forekommer spor- 

 vis i det saltsure Udtræk. 



I en anden Portion bestemtes Jernoxydul ved Titre- 

 ring med Kamæleon ; den samlede Jernmængde fandtes paa 

 samme Vis efter Beduktion med Zink. 



Det i Saltsyre uopløste Residuum udkogtes med en 

 Opløsning af kulsurt Natron til Bestemmelsen af Kiselsyren 

 i de dekomponerede Silicater, blev derpaa glødet til kon- 

 stant Vægt i Perrots Ovn og veiet. En Del deraf (0.8 

 til -0.9 Gr.) opsluttedes med kulsurt Kali-Xatron, Kisel- 

 syren fraskiltes, og af Filtratet udfældtes Lerjord -|- Jern 

 og Magnesia paa almindebg Maade. Det første Bundfald 

 opløstes efter Veiningen i koncentreret Saltsyre a , Jernet 

 bestemtes derpaa ved Fældning med Ammoniak. etterat 

 Lerjorden var fjernet med Katronlud. 



For med Hurtigbed at kunne bestemme Mængden af 



1 Hvad enten man anvender 100" eller 1)0", gjor ingen væsent- 

 lig Forskjel. 



: Ved denne Behandling' kan det være vanskeligt at faa Bund- 

 Faldet fuldstændig- opløst, mon dette i']' heller ikke nødvendigt, dadet 

 tilbageblevne uopløste vesentlig bestaar af Lerjord; selv om denne 

 skulde indeholde smaa .Mængder af Jern. er den ialfald ved Behand- 

 lingen med Saltsyre blevet skikket, til :it paavirkes af Natronlud, 

 hvorved Jernoxydet bliver uopløst. Dette har jeg ved Forsøg over- 

 bevist mig nm. 



water, to get rid of the salts. Tbis cannot however be 

 done entirely without boiling the mass ; but for ordinary 

 chemical analyses, repeated washing in cold water is quite 

 sufficient, the residue of salt still remaining in the sub- 

 stance being hardly appreciable. 



The portion of deposit vas then dried at about 100° 

 C. 1 , after which a few grammes, placed in a platina crucible, 

 were thoroughly heated in one of Perrofs furnaces, to de- 

 tei - mine the proportion of water and organic substances 

 (loss by ignition). From the same portion, 5 or 10 grammes 

 — according as the deposit was more or less homogeneous 

 in character — were next weighed off and boiled in a flask, 

 for about a quarter of an hour, along with 80"' or 40 cc of 

 diluted hydrocbloric acid (20 per cent). 



By exclusively adopting this method, fairly congruous 

 results may be obtained for the decomposable part of the 

 substance analysed, especially as regards the finer deposits 

 (Biloculina clay). With divers test-analyses specially made 

 to substantiate the trustworthiness of the-process in question, 

 the greatest difference did not exceed 2 or 3 per cent. 



After dissolving the sediment filtered from the decoc- 

 tion with hydrochloric acid, and diluting the fluid till its 

 volume liad reached 500'''' or 250"'. the iron, alumina, 

 lime, and magnesia in 50" or 25"' of the solution — 

 drawn off with a pipette — were determined in the usual 

 manner. 



Previous evaporation, to get rid of the silicic acid. 

 was had recourse to for some of the analyses, though such 

 is not strictly necessary, traces only of tbat constituent oc- 

 curring in the decoction. 



In another portion of the same sample, protoxide of 

 iron was determined by titrating with permanganate of 

 potash, and the total . amouut of iron in like manner 

 after desoxidation with zinc. 



The residue insoluble in hydrocbloric acid was boiled 

 along with a solution of carbonate of soda. to determine 

 the pro2rørtion of silicic acid present in the decomposed 

 silicates, and was tben thoroughly heated in one of Per- 

 rofs furnaces, and weighed. Part of it (0.8 — 0.9 gramme) 

 was fused with carbonate of potash and soda. the silicic 

 acid got rid of, and alumina, iron, and magnesia pre- 

 cipitated in the usual manner. The first deposit was dis- 

 solved, after weighing. in concentrated hydrocbloric acid, 2 

 and the iron tben present determined by precipitation with 

 ammonia, the alumina having been previously abstracted by 

 means of soda-lye. 



To determine in a short space of time the proportion 



1 Whether at 100° 'or 1 lo" makes very little difference. 



a Thus treated, it may lir dillieult to dissolve the whole of the 

 deposit; but this is not necessary, the umlissolved residue consisting 

 chictly of alumina; aud even if it should eontain minute quaiitities 

 of iron; the treatment with hydrochloric acid will have rendered 

 it susceptible to the action of soda-lye. causing the oxide of iron 

 to*remain umlissohcd. Oi' this I have satistied myself by direct 

 experiment. 



