Price 50 cents 
 
 REPRINT AND CIRCULAR SERIES 
 
 OF THE 
 
 NATIONAL RESEARCH 
 COUNCIL 
 
 RESEARCHES ON MODERN BRISANT NITRO EXPLOSIVES 
 BY C. F. VAN DUIN AND B. C. ROUTERS VAN LENNEP 
 
 TRANSLATED BY CHARLES E. MUNROE 
 
 Chairman, Committee on Explosives Investigations, Division of 
 
 Chemistry and Chemical Technology, National 
 
 Research Council 
 
 ^^^^M 
 
 California 
 Regional 
 'acility 
 
 Published in Recueil des Travaux Chimiques des Pays-Bas 
 February 15, 1920, vol. 39, no. 2, pp. 145-177
 
 Announcement Concerning Publications 
 
 of the 
 National Research Council 
 
 The Proceedings of the National Academy of Sciences 
 
 has been designated as the official organ of the National Research 
 Council for the publication of accounts of research, committee and 
 other reports, and minutes. 
 
 Subscription rate for the "Proceedings" is $5 per year. Business 
 address: Home Secretary, National Academy of Sciences, Smith- 
 sonian Institution, Washington, D. C. 
 
 The Bulletin of the National Research Council 
 
 presents contributions from the National Research Council, other 
 than proceedings, for which hitherto no appropriate agencies of 
 publication have existed. 
 
 The "Bulletin" is published at irregular intervals. The sub- 
 scription price, postpaid, is $5 per volume of approximately 500 
 pages. Numbers of the "Bulletin" are sold separately at prices 
 based upon the cost of manufacture (for list of bulletins see third 
 cover page). 
 
 The Reprint and Circular Series of the National Research 
 
 Council 
 
 renders available for purchase, at prices dependent upon the cost 
 of manufacture, papers published or printed by or for the National 
 Research Council (for list of reprints and circulars see third cover 
 page). 
 
 Orders for the "Bulletin" or the "Reprints and Circulars" of the 
 National Research Council, accompanied by remittance, should be 
 addressed: Publication Office, National Research Council, 1701 
 Massachusetts Avenue, Washington, D. C.
 
 REPRINT AND CIRCULAR SERIES 
 
 OF THE 
 
 NATIONAL RESEARCH COUNCIL 
 
 NUMBER 15 
 
 RESEARCHES ON MODERN BRISANT NITRO 
 EXPLOSIVES* 
 
 BY C. F. VAN DUIN AND B. C. ROETERS VAN LENNEP 
 
 Translated by CHARLES E. MUNROE 
 
 Chairman, Committee on Explosives Investigations, Division of Chemistry and Chemi- 
 cal Technology, National Research Council 
 
 INTRODUCTION 
 
 In the examination of explosives there are actually established 
 the following constants according to the methods developed by 
 Bichel, 1 Kast, 2 Will, Dautriche, Noble, etc., and to the principles 
 enunciated by Berthelot, 3 Sarrau, Vieille, Abel and others: 
 
 1. Velocity or rate of explosion 
 
 2. Energy liberated by the explosion 
 
 3. Composition of the products formed 
 
 4. Calculated temperature produced by the explosion 
 
 5. Pressure in a 15 1. Bichel 4 bomb 
 
 6. Enlargement of the cavity in a Trauzl lead block 
 
 7. Duration and length of the flame produced by the explosion 
 
 of 100 g. of the substance 
 
 8. Stability and the initiating explosion temperature 
 
 9. Sensibility to shock 
 
 The greatest weight ought undoubtedly to be assigned to the 
 constants 1, 2, 3, 8 and 9 ; but as we have not had at our disposal the 
 necessary devices with which to measure the first three, we have 
 
 * From Recueil des Travaux Chimiques des Pays-Bas, vol. 39, no. 2, pp. 145-177, 
 Feb. 15, 1920. 
 
 1 Bichel, "Methoden und Apparate der Sprengstoff A. G. Carbonit, zur Priifung von 
 Sprengstoffen . " 
 
 2 Kast, "Anleitung zur chemischen und physikalischen Untersuchung der Spreng- 
 und Ziindstoffe," Braunschweig, 1909. 
 
 3 Berthelot, "Sur la force des matieres explosives d'apres la thermochimie," Paris, 
 1883. 
 
 4 Bichel, Zeitschr.f. d. ges. Schiess- und Sprengstoffiv., 3, 365 (1901).
 
 2 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 been limited to the determination of the stability and of the sensi- 
 tiveness to shocks. The two last properties have had a special 
 significance since Fliirscheim 1 introduced into the technique of 
 explosives substances having very mobile nitro groups, such as 
 tetranitroaniline, while others have sought to employ the dini- 
 tramines. 2 
 
 To render the researches on strongly nitrated compounds having 
 one or more methylnitroamino groups or very mobile nitro groups 
 as complete as possible, we have also sought to derive from the 
 tetranitrophenylmethylnitramine, 2,4,6,2',3',4'-hexanitrodiphenyl- 
 amine and tetranitrophenol some of those bodies which have not 
 been employed as explosives because of their mode of prepara- 
 tion or their properties. 
 
 PREPARATION AND PROPERTIES 
 
 I. Tetranitrophenylmethylnitramine (Tetryl) 
 
 On the subject of this explosive one should consult 
 existing works. 3 
 
 N0 2 
 
 II. 2,3,4,6-Tetranitroaniline 
 
 To the communications which we have already made 4 on the 
 subject of the preparation and the properties of this compound, 
 we now add the results of the study of the action of moist acetone 
 upon it, in which we have identified the products formed in this 
 reaction. By the action of moist acetone on the tetranitroaniline 
 NH there are formed large quantities of hydrogen cyanide, 
 together with the trinitroamidophenol previously men- 
 tioned. By distillation in a vacuum on the water 
 bath, after having eliminated the acetone, there is ob- 
 tained an acid liquid, which after neutralization with 
 ammonia is colored black on warming after the addition of a 
 silver salt. 5 We expected to find formic acid present, but have 
 
 1 D. R. P. 241,697 and 243,079. 
 
 2 Brevet francais, 391,107. 
 
 3 Van Uuin, Rec. trav. (him., 37, 111 (1917). 
 
 4 Van Duin, loc. cit., 114. 
 
 5 This distillation ought to be conducted with great care, for there remains in the 
 flask very impure trinitroamidophenol, which can give rise to an explosion. See the 
 experience of Van Romburgh on the trinitromethylnitraminophenol. Van Romburgh, 
 Versl. Kon. Akad. van Wetensch. Amsterdam, 23, 1340 (1915).
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 3 
 
 been unable to identify it in the products of the reaction. Apart 
 from the water, which is a most important fraction, the distilled 
 liquid appears to be composed as follows: 
 
 a. Formaldehyde. This can be shown by a direct reduction of 
 a solution of ammoniacal silver oxide and also of a dilute solution 
 of potassium permanganate. It produced also an intense red color, 
 characteristic of formaldehyde, when the liquid was mixed with a 
 solution of 0.1 g. of morphine hydrochloride in 5 cc. of concentrated 
 sulphuric acid. 1 
 
 b. Acetic acid. After neutralization with soda the liquid was 
 evaporated and the salt obtained identified as sodium acetate, as 
 follows: ferric chloride produced a red color and on boiling there 
 was formed a precipitate of basic ferric acetate. On heating with 
 alcohol and some drops of concentrated sulphuric acid, the odor 
 of ethyl acetate was immediately perceived. On heating with a 
 solution of ^-nitrobenzyl bromide in 63% alcohol as indicated by 
 Emmet Reid, 2 crystals were obtained which fused at 78; when 
 these were mixed with ^-nitrobenzyl acetate, prepared according 
 to the method described by the same author, there was no lowering 
 of the melting point. 
 
 It was possible to show the presence of oxalic acid in the residue 
 remaining in the flask. For this purpose the residue was extracted 
 with hot water, neutralized by ammonia, and then acidified by 
 acetic acid. On addition of a solution of calcium chloride 
 there was observed the formation of a precipitate which immedi- 
 ately, when warm, reduced the solution of permanganate of 
 potash slightly acidified by sulphuric acid. Thus products formed 
 in the decomposition of aqueous solutions of di-isonitroso and 
 isonitroso acetone were indicated, though we were not able to 
 isolate them. 
 
 It is already known that when an aqueous solution of di-isonitroso 
 acetone is heated, it decomposes as follows: 
 
 H o H 
 / / / 
 
 -C > CO 2 + 2HCN + H 2 O 
 
 \ 
 NOH 
 
 At the same time there is produced a little ammonia, hydrogen 
 cyanide and oxalic acid : 3 
 
 1 Tendler and Manninck, Zeitsch. f. anal. Chem., 48, 310 (1909). 
 
 2 Emmet Reid, Journ. Amer. Chem. Soc., 39, 124 (1917). 
 
 a Von Pechmann and Wehsarg, Ber. d. deutsch. chem. Ges., 21, 2989 (1888).
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 COO NH 4 COOH + HCN 
 
 For the isonitroso acetone we have found that when this body is 
 heated to 140 in the presence of dilute hydrochloric acid, it de- 
 composes as follows: 1 
 
 OH HO CH 3 o 
 
 CH 3 C C + 2H 2 O - * NH 3 + C OH + C OH 
 
 \ 
 
 NOH 
 
 However, given the low temperature at which the action of moist 
 acetone on tetranitroaniline occurs, the reaction can be accurately 
 represented as follows: 
 
 OH H CH 3 
 
 / / II 
 
 CH 3 C C + 2H 2 O > NH 2 OH + C = O + C O 
 
 \ \ \ 
 
 NOH H OH 
 
 This is why it appears, in accordance with the facts cited above, 
 that the acid reaction of the distillate obtained by Van Romburgh, 
 through the action of moist methyl ethyl ketone on tetranitro- 
 phenylmethylnitramine, should be attributed to the presence of 
 acetic acid, and its reducing property to the presence of acet alde- 
 hyde. These compounds then appear to be produced from isonitro- 
 somethylethylketone, as follows: 
 CH 3 
 
 C = NOH ^O 
 
 I + 2H 2 O - > CHjCOOH + CHsCf + NH 2 OH 
 
 c = o NH 
 
 CH 3 
 
 It may be considered established that Fliirscheim was the first to 
 knowingly prepare the tetranitroaniline. However, we believe 
 that Witt and Witte 2 have had this substance in hand and have 
 considered it as a trinitroaniline. These authors state that they 
 obtained the 3,4-dinitroacetanilide, 3 melting point 144, when a 
 
 1 Treadwell and Steiger, Ber. d. deutsch. chem. Ges., 15, 1059 (1882). 
 
 2 Witt and Witte, Ibid., 41, 3095 (1908); see also Fliirscheim, 
 D. R. P. 243,079. 
 
 3 Wender, Gaz. chim. ital, 19, 225 (1889).
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 5 
 
 solution of 10 parts of m-nitroacetanilide in 100 of 100% sulphuric 
 acid was mixed below 5, with a solution of 15 parts of dry potas- 
 sium nitrate in 100 of 100% sulphuric acid and the whole allowed 
 to remain at the ordinary temperature during 24 hours. 
 
 It is alleged that with longer standing there is formed, little by 
 little, trinitroaniline. The authors have not given the fusion point 
 or made other tests but they note that an analysis for nitrogen 
 gave results corresponding with the calculated value for the trinitro- 
 aniline. If, however, we consider that the nitrogen contents of 
 tri- and of tetranitroanilines differ but little (24.56% and 25.65%) 
 one may realize that an examination of impure tetranitroaniline 
 could lead one to believe it to have the composition of the trinitro 
 product. 
 
 We have sought to prepare the trinitroaniline by the method 
 described by Witt and Witte and, while we have carefully endeavored 
 to follow the instructions given, we have never obtained trinitro- 
 aniline but always the tetranitro compound. This was identified 
 not only by the fact that it had the same melting point and that 
 when mixed with this substance the melting point was not lowered, 
 but also because on boiling in the presence of moist acetone it gave 
 2,4,6-trinitroamidophenol. It may also be remarked here that by 
 this process a very pure tetranitroaniline is immediately obtained. 
 
 III. Acetyltetranitroaniline 
 H O 
 
 / ^ 
 N CCH 3 
 
 NO 2 lNO 2 
 
 NO 2 
 
 We prepared this body according to Fliirscheim 1 by heating pure 
 tetranitroaniline with acetic anhydride and a little sulphuric acid 
 at a temperature not above 80. After crystallization in glacial 
 acetic acid the tetranitroaniline fuses, with decomposition, at 169- 
 170 , cor, 
 
 IV. 2,4,6-Trinitroamidophenol 
 NH 2 
 
 NO 
 
 NO 2 
 
 1 Flurscheim, D. R. P. 241,697, Chem. Zentr. Bl., 1, 184 (1912) ; see also Witt and Witte, 
 oc. cit., for the preparation of acetylpicramide.
 
 6 B RISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 Fliirscheim 1 has demonstrated that tetranitroaniline can easily be 
 transformed into trinitroamidophenol, even at the ordinary tem- 
 perature, if there is added to the acetone solution of it an aqueous 
 solution of sodium acetate. As we have previously shown, apropos 
 of tetranitroaniline, because of water being present with the acetone 
 as well as sodium acetate, the alkaline reaction of this latter ought 
 to accelerate the transformation. We always prepare the trinitro- 
 amidophenol by dissolving 100 g. of tetranitroaniline in 600 cc. of 
 acetone with the addition of a little water. As soon as the tetra- 
 nitro compound is dissolved, we distil off the acetone until the tetra- 
 nitrophenol crystallizes out; this is then dried, washed with cold 
 alcohol and then recrystallized, at first from alcohol and then from 
 benzene. Melting point 178-179, cor. 
 
 V. 2,4,6-Trinitroamidoanisol 
 NH 2 
 
 OCH 3 
 NO 2 
 
 This product has also been prepared (following Fliirscheim) by 
 boiling crude tetranitroaniline with absolute methyl alcohol. The 
 compound obtained is twice crystallized from methyl alcohol. To 
 obtain a pure product it is indispensable that the alcohol be absolute. 
 Melting point 131, cor. 
 
 VI. 2,4,6-Trinitroamidophenetol 
 NH 2 
 
 OC 2 H 5 
 NO 2 
 
 The method of preparation of this substance is similar to that f 
 the trinitroamidoanisol, except that ethyl alcohol is used. We 
 have, however, noticed that when prepared by the use of impure 
 tetranitroaniline very little of the trinitroamidophenetol is obtained, 
 and that that which is formed is also impure. The product is 
 strongly contaminated by a fibrous compound which, can be elimi- 
 nated by crystallization in alcohol-acetone; 2 but then the trinitro- 
 
 1 Fliirscheim, D. R. P., 243,079, Chem. Zenlr. Bl, 1, 620 (1912); Proc. Chem. Soc. 
 London, 26, 81. 
 
 2 Many compounds studied by us could be obtained pure with facility by dissolving 
 in warm acetone, then precipitating the compound by the addition of cold alcohol; we 
 employ for this procedure the simplified designation "cristallisation dans I'alcool-acelone."
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEf 7 
 
 amidophenetol is contaminated by trinitroamidophenol, which is 
 difficult to eliminate by crystallization. 1 While the purification of 
 tetranitroaniline by crystallization from glacial acetic acid is some- 
 what laborious, we recommend with some insistence the employ- 
 ment of the recrystallized product in the preparation of the trinitro- 
 amidophenetol. Melting point, 107, cor. 
 
 VII. 2,4,6-Trinitrometaphenylenediamine 
 
 N 
 
 NO 2 
 
 This compound was prepared by Nolting and Collin 2 from diethyl 
 ether, styphnic acid and ammonia in absolute alcohol. They have 
 described the compound as being yellow, crystalline, and fusing 
 only at a high temperature. Barr 3 proceeded as Nolting and Collin 
 did; he described the compound as yellow, fusing above 250. 
 Blanksma 4 prepared the body from 1,3, 2,4,6, -oxymethylchlortri- 
 nitrobenzene, but he does not state either the color or the melting 
 point. Meisenheimer and Patzig 5 obtained the compound by the 
 action of hydroxylamine on an alkaline alcoholic solution of sym- 
 metric trinitrobenzene ; the product principally formed was picram- 
 ide. Korner and Contardi 6 employed 2,4,6-trinitrodibrom and 
 dichlor benzenes with alcoholic ammonia. They obtained yellow 
 needles fusing above 250. Fliirscheim 7 employed tetranitroaniline 
 and ammonia; melting point, 275. Van Romburgh and Schepers, 8 
 by the prolonged action of a nearly saturated aqueous solution of 
 ammonia on 2,3,4,6-tetranitrophenylmethylnitramine, obtained the 
 trinitrometaphenylenediamine as a yellow-brown product fusing 
 above 270. 
 
 In our researches on the substitution of the methylnitroamino 
 group, 9 in the derivatives of tetranitrophenylmethylnitramine, we 
 
 1 Van Duin, Rec. trav. chim., 38, 89 (1919). 
 
 2 Nolting and Collin, Ber. d. deutsch. chem. Ges., 17, 260 (1884). 
 s Barr, Ibid., 21, 1546 (1888). 
 
 4 Blanksma, Rec. trav. chim., 21, 324 (1902). 
 
 5 Meisenheimer and Patzig, Ber. d. deutsch. chem. Ges., 39, 2533 (1906). 
 
 6 Korner and Contardi, Atti R. Accad. dei. Lincei Roma, 5, 17, 1, 465 (1908); 5, 18 
 1, 93 (1909); Chem. Zentr. Bl., 1908, II, 47; 1909, I, 1157. 
 
 7 Fliirscheim, loc. cit. 
 
 8 Van Romburgh and Schepers, Versl. Konin. Akad. van Wetensch. Amsterdam, 22, 
 293 (1913); Schepers, "Dissertatie," Utrecht, 1913, 41. 
 
 9 Van Duin, Rec. trav. chim., 38, 89 (1919).
 
 8 RRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 have obtained the trinitrometaphenylenediamine by the action of a 
 concentrated ammoniacal solution of 2,4,6-trinitroanilidophenyl- 
 methylnitramine. The product is dark brown and has a melting 
 point of 295, cor. (after 250, 10 per minute). After crystalli- 
 zation in glacial acetic acid, the melting point is raised to 301 , cor. 
 (after 250, 10 per minute). 
 
 We have also obtained this product by the action of ammonia 
 on the 2,4,6-trinitrodimethylaminophenylmethylnitramine; the 
 compound was dark brown and with a melting point of 290 , cor. 
 After crystallization in glacial acetic acid it rose to 296, cor. 
 (here also after 250, 10 per minute). In both cases the color of 
 the product was darkened by recrystallization. 
 
 We have obtained the purest trinitrometaphenylenediamine by 
 replacing a halogen atom or an oxyalkyl group by an amino group, 
 which has been achieved by the action of ammonia on trinitro- 
 amido-anisol and -phenetol. That an oxyalkyl group, placed near 
 a similar analogue, can be easily replaced by ammonia has been 
 already shown in the case of the isobutylic ether of the trinitro- 
 methylnitraminophenol. 1 Other examples show that this replace- 
 ment is general for the ethers of trinitroamidophenol and trinitro- 
 methylnitraminophenol. In this last case only has there been 
 obtained, by very prolonged action, the aminotrinitrophenylmethyl- 
 nitramine, on the side, which imparts a more or less dark brown 
 color to the trinitrometaphenylenediamine. For the preparation 
 of the trinitrometaphenylenediamine there is poured on 10 g. of 
 pure and finely pulverized trinitroamidoanisol or phenetol 80 cc. 
 of a concentrated aqueous solution of ammonia (s. p. 0.90) and the 
 whole left standing for 24 hours at ordinary temperature with fre- 
 quent agitation; at the end of this time the reaction is complete. 
 The product formed is filtered, freed from ammonia by washing, 
 and boiled with alcohol in order to finally eliminate all traces of the 
 anisol or phenetol which have escaped attack. The melting point 
 is then about 285, cor. After crystallization from glacial acetic 
 acid it is 287-288, cor. (after 250, 10 per minute). There is 
 thus obtained a trinitrometaphenylenediamine in beautiful yellow 
 crystals with metallic luster. 
 
 1 Van Duin, Koll. Zeitschr., 17, 123 (1915). The action of a concentrated ammoniacal 
 solution of the ethers of trinitroamidophenol and trinitromethylnitroaminophenol en- 
 ables us to prepare the corresponding alcohols very pure, which is quite important since 
 easily accessible alcohols do not give crystallizable ethers and they react with tetra- 
 nitroaniline or tetranitrophenylmethylnitramine to give the ether oxides. See Van 
 Romburgh and Schepers, loc. cit.
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 9 
 
 NITROGEN DETERMINATION 
 
 0.1342 g. of the product gave 33.1 cc. of nitrogen at 17 and 768 mm. 
 
 Nitrogen found: 28.78%. 
 
 Nitrogen calculated for the trinitrometaphenylenediamine : 28.82%. 
 
 In general it may be said that this trinitrometaphenylenediamine 
 is obtained of a more or less dark color when one or two amino 
 groups are introduced into the molecule by means of the deriva- 
 tives of the methylnitroamino group. It is remarkable that the 
 melting point is higher for the impure compounds ; in addition, the 
 recrystallization of the impure compound increases the melting 
 point and at the same time darkens the coloration; it is necessary 
 then to use glacial acetic acid for the recrystallization. The pure 
 product commences to melt without decomposition at 287-288, 
 cor. (after 250, 10 per minute) ; shortly after fusion small gaseous 
 bubbles appear in the clear yellow liquid ; this marks the com- 
 mencement of the decomposition. 
 
 It may be said that the impure product having a more elevated 
 melting point and at the same time a dark coloration, may be com- 
 pared with that which has been observed for the trinitroamido- 
 phenol obtained by replacement of the methylnitramino group in 
 the trinitromethylnitraminophenol by the action of ammonia 1 and 
 removal of the ammoniacal salt. The trinitroamidophenol is brown 
 and melts at 181, cor., instead of 178-179, cor. 
 
 VIII-XII. 2,3,4,6-Tetranitrophenylnitramine and its derivatives 
 CH S CH S CH 3 
 
 / / / 
 
 N NO, N NO 2 N NO 2 
 
 NO/^NOa NO2f\NO 2 NO/ \N0 2 
 
 1 JN0 2 \/ H I JOCHs 
 
 NO 2 N0 2 NO 2 
 
 CH 3 
 
 / 
 N NO 2 
 
 oc 2 H 5 NH, 
 
 NO 2 NO 2 
 
 In the preparation of these bodies, we have followed precisely the 
 directions given by Van Romburgh and Schepers. 2 The tetranitro- 
 
 1 Van Duin, Rec. trav. chim., 38, 89 (1919). 
 
 2 Van Romburgh, Ibid., 8, 273 (1889); Van Romburgh and Scheper, Versl. Kon. 
 Akad. v. Wetensch. Amsterdam, 22, 293 (1913); Schepers, "Dissertatie," Utrecht,1913.
 
 10 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 phenylmethylnitramine such as we have used for other experiments 
 was purified by dissolving in warm nitric acid, sp. gr. 1.49, to which 
 we have added a little concentrated sulphuric acid, in order to 
 protect the nitro group in position 3; 1 then, after filtration, we 
 have precipitated the compound with sulphuric acid and have 
 recrystallized in benzene and dried with phosphoric anhydride. 
 Melting point, 146-147, cor. 
 
 The trinitromethylnitraminophenol 2 was obtained by placing the 
 crystallized nitramine in a mixture of sulphuric and nitric acid and 
 heating these with water until solution was complete. This phenol 
 was then precipitated from the solution by chlorhydric acid, filtered, 
 dried in a desiccator over soda and then recrystallized in benzene. 
 Melting point, 183, cor. 
 
 The trinitromethylnitramino-anisol 3 and -phenetol 4 were pre- 
 pared by boiling the crude nitramine with absolute methyl or ethyl 
 alcohols and then by recrystallization with these alcohols. Melting 
 points: anisol, 96-97, cor.; phenetol, 98-99, cor. It is to be re- 
 marked that the action of these alcohols when cold is much more 
 rapid with the nitramine than with tetranitroaniline and this shows 
 that there is a much greater mobility of the nitro group in the 3 
 position; this is in accordance with the results of the tests for sta 
 bility. 
 
 For the preparation of the aminotrinitrophenylmethylnitramine 
 we have followed the method of Schepers 5 but have operated as 
 follows : 140 g. of tetranitrophenylmethylnitramine were introduced 
 little by little into 400 cc. of a 10% aqueous ammoniacal solution 
 during which the reaction vessel was cooled with cold water. After 
 allowing the reaction to take place for one hour, the mixture was 
 poured into 400 cc. of a 10% chlorhydric acid and then the amino- 
 trinitrophenylmethylnitramine formed was filtered. Yield, 120 g. 
 
 XIII. 2 ) 4,6,3',4',6'-Hexanitrodiphenylether 
 NO 2 NO 2 
 
 NO 2 NO 2 
 
 1 Van Romburgh and Schepers, he. cit. 
 * Ibid. 
 
 3 Ibid. 
 
 4 Ibid. 
 
 6 Schepers, loc. cit., 39; this method requires much time for the preparation o large 
 quantities; our preparation is analogous to it.
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 1 1 
 
 To purify the product it was at first boiled with alcohol and then 
 crystallized from alcohol-acetone. Melting point, 188, cor. 
 
 According to the patent of the Westphalisch-Anhaltische Sprengs- 
 toff A. G. 1 the above compound was prepared from the di-, tri-, 
 tetra- and pentanitrodiphenylethers, which had a nitro group in 
 one of the benzene nuclei in the 3 position. As a point of depart- 
 ure we found it best to take the nitrodiphenylether which could be 
 most easily obtained from 2,4-dinitrochlorobenzene or picryl chloride 
 and the w-nitrophenol. We have always made use of the 2,4,6,3'- 
 tetranitrodiphenylether which can easily be prepared by following 
 the instructions of the patent. 2 There is thus obtained a com- 
 pound which melts at 172, cor., after recrystallization from glacial 
 acetic acid. This tetranitrodiphenylether can be easily trans- 
 formed into 2,4,6,3',4',6 / -hexanitrodiphenylether according to the 
 instructions of the patent 3 if it is suspended in concentrated sulphuric 
 acid and the mixture is poured into three to four times its volume 
 of mixed acid (nitric acid, sp. gr. 1.49 and 10% oleum). This 
 mixture is heated to 100 and after some time the hexanitrodiphenyl- 
 ether separates out as a solid. If, however, 10 g. of the 2,4,6,3'- 
 tetranitrodiphenylether is suspended in 50 cc. of concentrated sul- 
 phuric acid, and this mixture is poured into 150 cc. of mixed acid, 
 composed of 25 cc. HNO 3 , sp. gr. 1.49, and 125 cc. of 10% oleum, 
 and then heated for an hour or two at 90 , there is obtained a product 
 which on recrystallization from acetic acid or toluene melts at 
 203-204, cor. This is the pentanitrodiphenylether and not the 
 hexa compound. This penta compound is the 2,4,6, 3 ',4 '-penta- 
 nitrodiphenylether which is obtained according to the patent at 
 110 from 3,4-dinitrophenylether and the sulphuric-nitric acid 
 mixture. 4 
 
 Analysis: 0.1578 g. of the substance gave 23.90 cc. N at 18 under 776 mm. 
 
 Nitrogen found: 17.78%. 
 
 Nitrogen calculated for pentanitrodiphenylether: 17.73%. 
 
 Even if 10 g. of tetranitrodiphenylether was dissolved in 40 cc. of 
 nitric acid, sp. gr. 1.49, and the mixture at 45 poured into the 
 mixed acid, formed from 75 cc. of nitric acid, sp. gr. 1.49, and 150 
 
 1 D. R. P. 281,053; Chem. Zentr. Bl., I, 74 (1915). As we have not been able to obtain 
 this patent because of political events, we have followed Escales, "Nitrosprengstoffe," 
 81. 
 
 2 Following Escales, "Nitrosprengstoffe," Leipzig, 1916, 82. 
 
 3 Escales, loc. cit. 
 
 4 D. R. P. 281,053; Chem. Zentr. Bl, 1915, I, 74.
 
 12 BRISANT NITRO EXPLOSIVES: VAN DUIN AND VAN LENNEP 
 
 cc. of 10% oleum, and the whole heated to 90 during an hour or 
 two, we still obtain the pentanitroderivative cited above. How- 
 ever we have obtained the 2,4,6,3 / ,4',6'-hexanitrodiphenylether by 
 using more nitric acid (sp. gr. 1.49,) and 30% oleum. The solution 
 of 10 g. of 2,4,6,3'-tetranitrodiphenylether in 40 cc. of nitric acid, 
 sp. gr. 1.49, is brought to 70 in a mixture of 90 cc. of nitric acid, 
 sp. gr. 1.49, and 170 cc. of 30% oleum, and the whole heated during 
 four hours to 90. There is thus obtained a white solid which melts 
 in a sluggish manner between 240 and 265; its recrystallization 
 is attended with great loss (50%); it is best to use nitric acid, 
 sp. gr. 1.49, at first, and then glacial acetic acid. Melting point, 
 278, cor. (after 200, 10 per minute). However, the purest 
 product begins to soften about 265 . 1 
 
 When the filtrate from the recrystallization in nitric acid, sp. gr. 
 1.49, is poured into water, a precipitate is obtained which melts in 
 a sluggish manner between 215 and 250; this was at first believed 
 to consist of the 2,4,6,3',4 / ,6 / -hexanitrodiphenylether, contaminated 
 with 2,4,6,3',4 / -pentanitrodiphenylether. But if this product is 
 nitrated as directed above, that is to say with nitric acid of 1.49 
 sp. gr. and 30% oleum during four hours at 90, the melting point 
 does not vary appreciably. This mixture is then probably composed 
 of 2,4,6,3',4 / ,6 / -and 2,4,6,2',3',4'-hexanitrodiphenylether. 2 A de- 
 termination of nitrogen shows that this mass is composed almost 
 entirely of hexanitro derivatives. 0.1462 g. of the substance gives 
 24.20 cc. nitrogen at 22 under 767 mm. 
 
 Nitrogen found: 18.82%. 
 
 Nitrogen calculated for pentanitrodiphenylether : 17.73%. 
 
 Nitrogen calculated for hexanitrodiphenylether : 19.10%. 
 
 We have not tried to isolate from the mixture the hexanitro 
 derivatives, but hope to do this later. 
 
 As it appeared important to be able to compare the stabilities, 
 sensitiveness, etc., of 2,4,6,3 / ,4 / ,6'-hexanitrodiphenylether and 
 2,4, 6, 2', 4 ',6 '-hexanitrodiphenylether, we have attempted to pre- 
 pare this dipicrylic ether as follows: 
 
 At the outset we may say that we are not much surprised that we 
 did not obtain the result, inasmuch as the Griesheim Chemical 
 
 1 According to D. R. P. 281,053, already cited, the melting point is about 269, not 
 cor. 
 
 2 We do not include the formation of 2,4, 6, 2 ',3 ',6 '-hexanitrodiphenylether made by 
 us in our tests long ago for the preparation of 2,4,6,2',4',6'-hexanitrodiphenylether.
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 13 
 
 Company patented 1 the 2,4,6,2 ',4'-pentanitrodiphenylether but not 
 the dipicrylic ether, although the latter would be, as an explosive, 
 superior to the former. 
 
 1. Ten g. 2,4,6,2',4'-pentanitrodiphenylether were dissolved in 
 40 cc. of nitric acid, sp. gr. 1.49, and then nitrated during four hours 
 at 90 with a mixture of 90 cc. of nitric acid, sp. gr. 1.49, and 170 
 cc. of 30% oleum. The initial compound was found unaltered. 
 
 2. Two g. of 2,4,6,2',4'-pentanitrodiphenylether were boiled 
 during one-half hour with a mixture of 40 cc. of real nitric acid and 
 20 cc. of 20% oleum, after which the nitric acid was eliminated. 2 
 On pouring this mixture into water, no precipitate was obtained, 
 not even a cloud. The penta derivative must then have been oxi- 
 dized. 
 
 3. Nitration with real nitric acid and phosphoric anhydride. 
 On nitrating 2,4,6,2',4'-pentanitrodiphenylether with 30 cc. of real 
 nitric acid and 21 g. of P 2 Oi during two hours at 40 , 3 the initial 
 product remains unchanged. 
 
 4. It is the same if 3 g. of 2,4,6,2',4'-pentanitrodiphenylether are 
 treated with a mixture of 30 cc. of real nitric acid and 30 cc. acetic 
 anhydride, and heated for some time at 30-40, or allowed to re- 
 main for a week at ordinary temperatures; not the least change is 
 observed. 4 
 
 5. If a solution of picryl chloride or bromide 5 is boiled during 
 five hours with silver picrate, 6 there is no transformation. 
 
 6. The same result is obtained when 10 g. of picryl chloride are 
 heated seven hours at 100 with 2 g. of silver picrate. 
 
 7. A reaction is obtained, however, if given quantities of picryl 
 chloride and silver picrate are heated in absolute alcohol, but this 
 reaction is not of the kind desired. We dissolved 6.8 g. of 
 silver picrate in 230 cc. of boiling absolute alcohol, and then 
 added 5 g. of picryl chloride when the silver chloride commenced 
 immediately to separate. The whole was allowed to boil during 
 ten hours, after which the reaction was apparently complete. The 
 
 1 D. R. P. 81,970; Escales, "Nitrosprengstoffe," Leipzig, 1915, 67. 
 
 2 Picryl bromide appears to have been prepared from 2,4,-dinitrobromobenzene by 
 Jackson and Earle, Amer. Chem. Journ., 29, 212 (1903). 
 
 3 We do not nitrate at a higher temperature because already at 45 a notable amount 
 of N2Os from decomposition is formed. 
 
 4 See Franchimont and Friedmann, Rec. trav. chim., 28, 192 (1908). 
 
 5 Jackson and Earle, loc. cit. 
 
 6 Kast, Zeitschr. f. d. ges. Schiess- und, Sprengstoffw., 6, 32 (1911).
 
 14 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 silver chloride was filtered and the alcohol distilled. The residue 
 obtained was then extracted twice with water as rapidly as possible 1 
 in order to eliminate the silver picrate which had not been trans- 
 formed when the material was recrystallized in absolute alcohol. 
 After several recrystallizations there was obtained a compound 
 melting at 79 which is not lowered by mixing with 2,4,6-trinitro- 
 phenetol. The silver picrate has then acted to remove the chlor- 
 hydric acid produced from the action of picryl chloride on alcohol. 
 We believe that according to the results which have been enumerated 
 we are warranted in concluding that the formation of 2,4, 6,2', 4', 6'- 
 is impossible because of stereochemical reasons. 
 
 XIV. 2,4,6,2',4',6'-Hexanitrodiphenylsulphide 
 
 NO 2 NO 2 
 
 v / \ 
 
 "> s \ y N 2 
 
 "NO 2 NO 2 
 
 After having stated 2 that the nitro derivatives of diphenylsulphide 
 can be prepared by the action of alkaline sulphides on chloronitro- 
 benzene, "Sprengstoff A. G. Carbonit" 3 found it could obtain a 
 better yield of the 2,4,2', 4'-tetranitrodiphenylsulphide by the action 
 of two molecules of 2,4-dinitrochlorobenzene and two molecules 
 of sodium hyposulphite, according to the following expression : 
 
 2C 6 H 3 (NO 2 ) 2 C1 + 2Na 2 S2O 3 = C 6 H 3 (NO 2 ) 2 S C 6 H 3 (NO 2 ) 2 + 2NaCl + Na 2 S 3 O 6 (I) 
 
 When, however, they sought to obtain the dipicrylsulphide from 
 picryl chloride by a similar reaction they obtained a dark flocculent 
 mass. From the newer researches of the "Sprengstoff A. G. Car- 
 bonit," 4 the 2,4,6,2 / ,4 / ,6'-hexanitrodiphenylsulphide can be obtained 
 provided but one molecule instead of two of the hyposulphite is 
 employed and the sulphuric acid formed in the reaction is neutralized. 
 The expression for this reaction is : 
 
 2C 6 H 2 (NO 2 ) 3 C1 + Na 2 S2O 3 + H 2 O = C6H 2 (NO 2 ) 3 .S.C 6 H 2 (NO 2 ) 3 +2NaCl + H 2 SO 4 (II) 
 
 The yield is about 90%. It is necessary to employ the least 
 excess of hyposulphite in the reaction as decomposition takes place 
 at the same time the color is darkened. The directions of the 
 patent for the formation of hexanitrodiphenylsulphide are precise; 
 
 1 This is for the removal of the trinitrophenetol of hydrolysis but even with boiling 
 water sufficient of the product remains for identification. 
 
 2 Willgerodt, Beilstein II, 1896, 803. 
 
 3 D. R. P. 94,077. 
 
 4 D. R. P. 275,037; Chem. Zentr. Bl., 1914, II, 97. Escales, "Nitrosprengstoffe," 79, 
 197.
 
 BRISANT NITRO EXPLOSIVES: VAN DUIN AND VAN LENNEP 15 
 
 we have never observed decomposition when using pure picryl 
 chloride and the theoretical quantity of hyposulphite; the least 
 excess of hyposulphite, on the contrary, produces decomposition. 
 If one wishes to obtain a crude product which is sufficiently purified 
 by a single recrystallization from alcohol-acetone, it will be prudent 
 to make use of hyposulphite in a quantity a little below the theo- 
 retical value. 
 
 We have prepared the hexanitrodiphenylsulphide as follows: 
 after the patent of which we have spoken a solution of 20 g. of picryl 
 chloride in 100 cc. of 95% alcohol with 3.8 g. of MgCO 3 suspended 
 in the solution is boiled in a flask furnished with a reflux condenser. 
 There is then added, little by little, with continual agitation, 10 g. 
 of finely pulverized Na 2 S 2 O3.5H 2 O (theory 10.03 g.). New portions 
 of the hyposulphite are added when that which has been introduced 
 has completely reacted. When the reaction is finished the filtered 
 crude product is washed with alcohol, then weak chlorhydric acid, 
 and then water. The yields were 14.45 and 14.65 g. of the hexanitro 
 product, which corresponds to 78.4 and 79.5%. As the patent 
 claims a yield of =>=90% we have tried the reaction with greater 
 quantities of picryl chloride, and using 60 g. of this substance the 
 yield of hexanitrodiphenylsulphide fell to 40 g. or 72.4%. This is 
 why after establishing the fact that the quantity of sulphuric acid 
 corresponds to the weight of hexanitrodiphenylsulphide we have 
 determined the quantity of acid in the filtrate of a preparation 
 made from 20 g. of picryl chloride. From reaction II it is shown 
 that 3.963 g. of sulphuric acid should be formed; we have found 
 3.721 and 3.725, which corresponds to 93.9% of the theory. In 
 this trial we have anew obtained 79.4% of hexanitrodiphenyl- 
 sulphide (see above) which ought to produce more sulphuric acid 
 as indicated by reaction II. The appearance of a disagreeable odor 
 of an organic sulphur compound 1 argues also in favor of this inter- 
 pretation; however, we are not able to indicate the nature of this 
 secondary reaction. The hexanitrodiphenylsulphide is best re- 
 crystallized in alcohol-acetone or in glacial acetic acid; the purest 
 product has a melting point of 233-234, cor. 2 
 
 XV. 2,4,6,2 ',4 ',6 '-Hexanitrodiphenylsulphone 
 NO 2 O.x ^O NO 2 
 
 NO 2 < > S <( >NO 2 
 
 NO 2 
 
 1 The hexanitrodiphenylsulphide has no odor. 
 
 2 According to Beilstein II, 803, the melting point is 226, not cor.
 
 1 6 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 Following D. R. P. 269,862 1 of the "Sprengstoff A. G. Carbonit," 
 hexanitrodiphenylsulphide can be transformed into its corresponding 
 sulphone by treating it in suspension in nitric acid, or in a mixture 
 of chromic and acetic acids, with bichromates or oxidizing com- 
 pounds of manganese. This compound has been assigned a per- 
 oxide formula: 
 
 C,H,(NO,)3v /O 
 
 \o/ I 
 
 /*\ \ 
 
 C 6 H 2 (N0 2 )3 / X 
 
 but this has been contested by Stettbacher 2 on account of the great 
 stability of the compound. Up to the present we have obtained 
 no other hexanitro derivative of oxidation, following the process 
 indicated, on using other oxidizers such as CrO 3 and K 2 Cr 2 O 7 . The 
 treatment of the sulphide by potassium permanganate in suspension 
 in nitric or sulphuric acid leaves the material unchanged. 
 
 The use of the theoretical quantity of chromic acid having given 
 us satisfactory results, we operated in the following manner: 17.1 g. 
 of hexanitrodiphenylsulphide were put in suspension in 100 cc. of 
 nitric acid, sp. gr. 1.49, then 5 g. of finely pulverized chromic acid 
 were added. 3 The mass was allowed to stand at the ordinary tem- 
 perature, with agitation from time to time, until the disappearance 
 of the chromic acid, which took about 15 days. During this time 
 the sulphide disappeared also, while there was formed a voluminous, 
 very slightly colored product. This was filtered, washed until acid 
 reaction disappeared, recrystallized in alcohol-acetone, and the 
 sulphone obtained. This was a yellowish white compound, melting 
 at 307, cor. (after 250, 10 per minute). 4 
 
 Analysis: 0.127 g. of substance gives 18.75 cc. nitrogen at 16 and 772 mm. 
 
 Nitrogen found: 17.40%. 
 
 Nitrogen calculated for hexanitrodiphenylsulphone: 17.22%. 
 
 By this process there is obtained a sulphone contaminated by a 
 little sulphide which may be removed by solution in nitric acid of 
 1.49 sp. gr. In fact, it is necessary to employ as much as 200 cc. 
 of nitric acid for 10 g. of sulphide. On account of the slowness of 
 the reaction we have tried to increase its activity by heat, but we 
 nave not obtained any better result with either nitric or acetic acid. 
 
 1 Escales, "Nitrosprengstoffe," Leipzig, 1915, 81, 283. 
 
 2 Stettbacher, Zeitschr. f. d. ges. Schiess- und Sprengstoffw., 11, 115 (1916). 
 
 3 It is strongly recommended that finely pulverized chromic acid be used because 
 the reaction is very slow. 
 
 4 F.-om this melting point, the explosion temperature of 250-255 given in the patent 
 must be inexact.
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 1 7 
 
 We entirely agree with the statement of Stettbacher 1 as concerns 
 the constitution of the product. To the argument furnished by this 
 author as to resistance to high temperatures, we wish to add: 1. 
 No iodine is separated when the sulphone, dissolved in glacial acetic 
 acid, is treated with potassium iodide. 2. Sensitiveness to shock: 
 as measured by the drop hammer it is less sensitive than the 
 sulphide 2 ; if the sulphone were a peroxide, its sensitiveness would 
 increase. 
 
 XVI. 2,4,6-Trinitrodimethyldinitraminebenzene 
 CH 3 NO 2 
 \/ 
 
 NO; 
 
 NO 2 
 
 This compound has been prepared by Van Romburgh from the 
 1,3-dimethylphenylenediamine, 3 the tetramethylmetaphenylenedi- 
 amine 4 and the 2,4,6-trinitro-3-methylaminophenylmethylnitranr 
 ine; 5 by Blanksma 6 from the l,3-dimethylamine-2,4-dinitrobenzene; 
 and by Blanksma and Meerum Terwogt 7 from the 4,6-dinitrodi- 
 methylmetaphenylenediamine with concentrated nitric acid. 
 
 We made use of a process analogous to that of Van Romburgh, 8 
 with the difference that we have not used as our point of departure 
 the monomethyl compound but have employed the dimethylamino- 
 trinitrophenylmethylnitramine which we have obtained from the 
 tetranitrophenylmethylnitramine by using less of the dimethyl- 
 amine than Schepers. 9 This last preparation was made as follows: 
 To 300 cc. of the 8% aqueous solution of dimethylamine was added, 
 in small portions, with constant cooling by water, 140 g. of tetra- 
 nitrophenylmethylnitramine and the mixture allowed to stand for 
 half an hour. It was then poured promptly into 250 cc. of 10% 
 chlorhydric acid and the dimethylaminotrinitrophenylmethylni- 
 
 1 Stettbacher, loc. cit. 
 
 2 See page 29, Table II. 
 
 3 Van Romburgh, Rec. trav. chim., 6, 251 (1887). 
 
 4 Van Romburgh, Ibid., 7, 3 (1888). 
 
 5 Van Romburgh, Ibid., 8, 280 (1889). 
 
 6 Blanksma, Ibid., 27, 27, 49 (1907). 
 
 i Blanksma with Meerum Terwogt, Ibid., 21, 291 (1901). 
 
 8 Van Romburgh, Ibid., 8, 280 (1889). 
 
 9 Schepers, "Dissertatie," Utrecht, 1913, 48; used for 3 gr. of nitramine, 5 gr. dimethyl-
 
 18 BRISA NT NITRO EXPLOSIVES: VAN D UIN AND VAN LENNEP 
 
 tramine formed was filtered, washed until the disappearance of the 
 acid reaction, and then boiled with alcohol. Yield, 100 g. 1 The 
 crude product was then introduced in small portions into nitric acid, 
 sp. gr. 1.49, and heated to 70-80, in order to effect the replace- 
 ment of the dimethylamine group by the methylnitramino group. 
 If sufficient nitric acid is employed there is obtained by recrystalli- 
 zation a very nearly colorless, almost white, product. This crystal- 
 lization, as stated by Van Romburgh, is best effected in alcohol- 
 acetone. The purified product melts at about 206, cor. (after 
 150, 10 per minute); it is to be remarked that the technical 
 preparation corresponds to that of Van Romburgh who used al- 
 kalized metaphenylenediamines. 2 
 
 XVII and XVIII. Dipicrylamine and 2,4,6,2',3',4'-Hexanitrodiphenylamine 
 NO 2 NO 2 NO 2 NO 2 NO 2 
 
 >NO 2 N0 2 <T > N ^ >NO 2 
 
 NO 2 NO 2 NO, 
 
 For the preparation and properties of these bodies, consult the 
 published memoirs. 3 
 
 xix. Tetra- We have prepared tetranitrophenol according to 
 
 nitrophenol Blanksma 4 from w-nitrophenol and sulphuric-nitric 
 
 acid. The use of real nitric acid is not essential; 
 
 NO:/ x iNO 2 w ith acid of sp. gr. 1.49 the reaction goes on as described 
 
 by Blanksma. However, it is well not to use at one 
 
 N 2 time more than 5 g. of m-nitrophenol. By recrystal- 
 
 lization from dry chloroform we have obtained a melting point of 
 
 140, cor. 
 
 STABILITY 
 
 Although the literature of chemistry is quite extended on the 
 subject of the stability of nitroglycerine, guncotton and smokeless 
 powder, it is, on the contrary, very limited as concerns the brisant 
 military nitro explosives and explosives used for industrial purposes. 
 For these last, the stability constants have been established from 
 preliminary trials by the Eighth International Congress of Applied 
 Chemistry 5 on the initiative of Will and Lenze. The German 
 
 1 See Van Dum, Rec. trav. Mm., 37, 111 (1917). 
 
 2 French patent 391,107; Escales, "Nitrosprengstoffe," Leipzig, 1915, 282. 
 
 3 Van Duin and Roeters Van Lennep, Rec. trav. Mm., 38, 358 (1919). 
 Blanksma, Ibid., 21, 254 (1902). 
 
 5 Zeitschr. j. d. ges. Schiess- und Sprengstoffw., 7, 435 (1912); 8, 345, 370 (1913).
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 1 9 
 
 Railway Commission has adopted these constants for nitro ex- 
 plosives. 1 The reason for much of the delay in the examination 
 of nitro explosives is due to the peculiar characteristics of the nitro 
 compounds, picric acid and trinitrotoluene. Verola 2 has shown 
 that trinitrotoluene may be maintained during 100 hours at 100 
 without its showing any alteration; the setting point, 80.7, does not 
 change. We have no similar data for picric acid at our disposal; 
 we notice only that in Holland picric acid (and the same may be 
 said of trinitrotoluene) must respond to such a trial as the following : 
 10 g. heated at 95 during three hours did not give off the least 
 trace of nitrogen oxides. Picric acid endures this test very well; 
 it may even be repeated during several weeks by heating it each 
 day during eight hours to 95, without the least nitrous vapor 
 making its appearance. 
 
 But the question of the stability of brisant nitro explosives en- 
 tered on a new phase with the employment of nitro compounds con- 
 taining one or more methylriitramino or mobile nitro groups. Of 
 the first group, that of the nitramines, it is not necessary to again 
 take count. The trinitrophenylmethylnitramine, used nearly ex- 
 clusively at the present time, is very stable when in a sufficiently 
 pure condition. It is quite otherwise with tetranitrophenylmethyl- 
 nitramine and tetranitroaniline, which have been described as very 
 stable, 3 they being, however, rapidly decomposed even if they are 
 very pure. 4 This is the more remarkable if we recall that tetra- 
 nitroaniline is more stable at 70 than nitroglycerine, while at the 
 ordinary temperature the contrary is true. The Nobel Dynamite 
 Factory at Avigliana preserves today 200 g. of the nitroglycerine 
 which was prepared for the first time in 1848 by Sobrero; the ni- 
 trogen content of this specimen has not varied up to the present, 
 nor has it shown the least trace of acid. 5 Tetranitroaniline, on the 
 contrary, resists well tests of stability at 71 6 while in different 
 
 1 Escales, "Ammonsalpetersprengstoffe," Leipzig, 1909, 120. 
 
 2 Verola, Zeitschr. f. d. ges. Schiess- und Sprengstoffw., 7,211 (1912); Escales, "Nitro- 
 sprengstoffe," Leipzig, 1915, 297. 
 
 3 Fliirscheim, Zeitschr. f. d. ges. Schiess- und Sprengstoffw., 8, 185 (1913). 
 
 4 The communication of Claessen in Holland patent no. 5,240, stating that the 
 2,3,5,6-tetranitro-anisol is notably more stable than the trinitrophenylmethylnitramine, 
 ought to be regarded as inexact. This conclusion was based solely on the high tem- 
 perature of inflammation, 300. 
 
 5 Escales, "Nitroglycerine and dynamite," Leipzig, 1908, 3. 
 
 6 Kast, "Spreng- und Ziindmittel," 1909, 952.
 
 20 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 countries a well determined decomposition has been observed at 
 ordinary temperatures after 9 weeks. 
 
 In our researches on the stability of brisant nitro explosives, we 
 have not determined whether these explosives respond to the tests 
 of stability employed for safety explosives. These tests are of 
 value for technical products, but the Abel test, in which is observed 
 the first disengagement of nitrous vapors, is too sensitive for these 
 technical products. 1 This test is made in duplicate: 10 g. of the 
 explosive are heated in flasks of determined dimensions during 48 
 hours at 75, after which there should not be produced the least 
 sign of decomposition as shown by the appearance or odor. 2 For 
 nitro compounds, it is further stipulated if changes are produced 
 the sensitiveness to shock should be determined on a sample in 
 which the change is most marked. 3 As we have tried chemically 
 pure compounds only, we have observed, following Abel, whether 
 or not a change was produced by them in starch-zinc-iodide papers 
 moistened with glycerine solution. The eventual presence of 
 nitrogen oxides thus shown can be due only to the decomposition 
 of the explosive because of the purity of the same, and this gives 
 a good idea of the stability. In our stability trials we have first 
 sought to ascertain if the explosive responds to the test applied in 
 Holland to picric acid and to trinitrotoluene, that is to say, if on 
 heating 10 g. to 95 during 3 hours the least decomposition is to be 
 observed. If the result of this test was good, the substance was 
 then heated during 30 days at the rate of 8 hours per day. If after 
 this period the material showed no change, the test was stopped 
 since its prolongation seemed to be without significance. The 
 term of 30 days was fixed upon arbitrarily. If the explosive failed 
 to pass that test, a determination of its stability was made at lower 
 temperatures. 
 
 In the tests of stability at 32, 50, 95, 10 g. of the substance 
 were heated in flasks of the kind specially used for determining the 
 stability of smokeless powder and guncotton. These flasks have 
 a capacity of 29.3 cc. (=*=3 cc.) and a height measured just to the 
 stopper of 46.1 mm. (2); they ought to pass the test called the 
 "ether test," which consists in placing 10 cc. of ether in a flask and 
 weighing the whole; after 24 hours of rest at the ordinary tem- 
 
 1 Lenze, Zeitschr. f. d. ges. Schiess- und Sprengstoffw., 4, 303 (1909). 
 
 2 Zeitschr. f. d. ges. Schiess- und Sprengstoffw., 8, 370 (1913). 
 
 3 Escales, "Ammonsalpetersprengstoffe," 120.
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 2 1 
 
 perature the diminution in weight ought not to surpass 100 mg. 
 For the trials at other temperatures we have employed tubes 16 
 cm. in length and 17 mm. in diameter, with a 2 cm., well ground, 
 glass stopper; these tubes passed the ether test; they were placed 
 in water baths at constant temperature, which is not possible with 
 flasks. The zinc-iodide-starch paper was prepared as follows: 
 100 cc. of water containing 4 g. of starch and 20 g. of zinc chloride 
 were boiled during 2 hours. 2 g. of zinc iodide and a liter of water 
 were then added and the whole filtered. In this solution strips 
 of Schleicher and Schull's No. 590 filter paper were soaked during 
 one minute and then dried in darkness and preserved in hermeti- 
 cally closed boxes. For use the paper thus prepared is moistened 
 with a 10% solution of glycerine by placing a drop on the lower 
 part. We have described the preparation of this zinc-iodide- 
 starch paper because there is attributed to it a certain importance 1 
 and this is an essential thing for the comparison of these tests. 
 We give below results obtained with the different explosives and 
 data of the literature, the compounds, prepared according to the 
 methods described in the preceding pages, being in a state of great 
 purity. 2 
 
 1 . The following compounds have responded to the tests described 
 in the preceding pages, which consist in heating 10 g. of the ma- 
 terial to 95 without the least trace of nitrous acid having been 
 disclosed. These are: Trinitrophenylmethylnitramine, trinitro- 
 amidophenol, -anisol and -phenetol, trinitrometaphenylenediamine, 
 trinitromethylnitraminophenol, aminotrinitrophenylmethylnitram- 
 ine, hexanitrodiphenyl sulphide and sulphone, and dipicrylamine. 
 For each of these substances the test has been continued during 
 30 days at a rate of 8 hours per day without the least change. On 
 account of the grat importance of the trinitrophenylmethylnitram- 
 ine to our modern explosives industry, 3 we have continued the test 
 for this last during 185 days, without having observed the least 
 decomposition. In the case of the trmitrodimethyldinitramino- 
 benzene, one flask has shown a reaction after three days, while 
 after four days four flasks have shown a decomposition. 
 
 2a. Tetranitroaniline. According to Fliirscheim, 4 it is easy to 
 
 1 See Zeitschriftf. d. ges. Schiess- und Sprengstoffw., 9, 293 (1914); there is given, with 
 details, the preparation of the paper employed in England. 
 
 2 For more details, see Van Duin, "Dissertatie," Utrecht, 1918, 76-84. 
 
 3 See Van Duin and Brackmann, Client. Weekbl., 16, 501 (1919). 
 
 4 Flurscheim, Zeitschr. f. d. ges. Schiess- und Sprengstoffw., 8, 185 (1913).
 
 22 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 prepare a product which will not cause coloration of the reagent 
 paper after 60 minutes, exposure at 71; the same author indicates 
 that the stability at 120 equals that of tetryl. However, Stett- 
 bacher 1 has already shown that the purest tetranitroaniline de- 
 composes after 8 days, this being attributed to the reaction between 
 the orthonitro and the amino groups. 2 
 
 Test at 70. After two hours' heating, no trace of nitrogen oxide 
 was observed; this is also the case after three hours. 
 
 Test at 60. No decomposition after 18 hours, nor even after 
 26 hours. 
 
 At 50. We have heated the specimens day and night; the sev- 
 enteenth day two out of four flasks gave a positive reaction. 
 
 At 32. The reagent paper has shown the presence of nitrogen 
 oxide compounds after 9 weeks. 
 
 2b. Acetyltetranitroaniline. Special stability tests have not 
 been made on this. We mention it only because the product, 
 purified as described above, shows an acid decomposition after two 
 months of preservation in a well stoppered flask. 
 
 2c. Tetranitrophenylmethylnitramine. Stettbacher 3 has pre- 
 sented the hypothesis that this compound would be more stable 
 than tetranitroaniline. As will be shown, precisely the contrary is 
 true. 
 
 Stability at 60. Of four flasks tested, two have given after one 
 hour a feeble reaction with the zinc-iodide-starch paper; after four 
 hours the four flasks have each given a strong reaction. 
 
 Stability at 50. After 15 days the four samples have given a 
 decided reaction. 
 
 Stability at 32. Two flasks have given a positive reaction after 
 four weeks and the others after five weeks. 
 
 We insist that it is not advisable to introduce the reagent paper 
 into the tubes used for the stability test of tetranitro compounds, 
 which easily react when warmed with water vapor. This is why 
 we have made our tests in this search of the nitrogen oxide com- 
 pounds by introducing zinc-iodide-starch paper into the test tubes 
 from time to time and keeping them there for but about 30 seconds. 
 The same error attends the use of other reagents which have been 
 
 1 Stettbacher, Zeitschr f. d. ges. Schiess- und Sprengstoffw., 11, 114 (1916). 
 1 Stettbacher, loc. cit.; Escales and Stettbacher, "Initialexplosivstoffe," Leipzig, 1917, 
 250, 2,54. 
 
 3 Stettbacher, Zeitschr. f. d. ges. Schiess- und Sprengstoffw., 11, 114 (1916).
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 23 
 
 proposed for testing the stability of compounds easily yielding 
 nitrogen oxide derivatives in the presence of water vapor. 1 
 
 2d. TetranitrophenoL We have not determined the stability 
 of this compound except at 70, on account of the difficulty in ob- 
 taining the well purified product. The results obtained are other- 
 wise sufficient for all our purposes. The tetranitrophenol appears 
 to behave like other tetranitro compounds. The presence of 
 nitrogen oxides was not shown, however, after maintaining the 
 tetranitrophenol at 70 during three hours, but after four hours the 
 reaction was positive. 
 
 2e. 2,4,6,3' ,4' ,6'-Hexanitrodiphenyleiher. In spite of the pres- 
 ence of the mobile nitro group, no decomposition had taken place 
 after heating for three hours at 95 ; after eight hours the reaction 
 was positive. 
 
 2f. 2,4,6,2' ,3' ,4'-Hexanitrodiphenylamine. As this product is 
 difficult to obtain, we have simply observed whether its stability is 
 markedly diminished by the replacement of an immobile nitro 
 group in the dipicrylamine by a mobile group. It is in effect thus: 
 A strong reaction with the zinc-iodide-starch paper was observed 
 after heating one hour at 95. 
 
 3. Trinitromeihylnitramino-anisol and -phenetol. The stability 
 of this compound is markedly less than that of the corresponding 
 phenols; we have definitely proved our several preparations. The 
 anisol after heating for one hour at 90 gave no reaction, but after 
 two hours the reaction was positive. At 80 the test is negative 
 during 24 hours, positive after 26 hours. On heating the flasks to 
 70 at the rate of eight hours per day, a positive reaction was ob- 
 served after 15 days. At 50 at constant heat, 20 weeks were 
 needed to render the reaction positive. With phenetol, there was 
 no reaction after one hour's heating at 90, but a marked reaction 
 after two hours. The reaction is much more manifest than that of 
 anisol. At 30, a reaction after 12 hours (none after 10); at 70, 
 reaction after 7 hours' heating, at the rate of 8 hours per day; and 
 at 50, after 20 weeks. It should be noted that we have observed 
 a less stability for the phenetol among our preparations. 
 
 We ought also to observe that the nitrogen oxides produced in 
 our tests of stability of tetranitro compounds come from the 
 
 1 See Guttmann, Zeitschr. f. angew. Chem., 1897, 265; Egerton, Zeitschr. f. d. ges. 
 Schiess- und Sprengstorfiv., 8, 390 (1913). These authors have worked with solutions 
 of diphenylamines, especially naphthylamine and dimethylaniline, which also contain 
 water.
 
 24 BRISANT NITRO EXPLOSIVES: VAN DUIN AND VAN LENNEP 
 
 compounds themselves, and not from the reaction between the nitro 
 product and the moisture of the paper. This is clearly shown by 
 the test made with three tubes each containing 10 g. of tetranitro- 
 aniline heated in a water bath at 70. The first tube was tested 
 after one hour and it gave no positive reaction ; after two hours the 
 second tube was also examined; neither of the two tubes gave a 
 positive reaction, but after three hours all three tubes gave a posi- 
 tive reaction. It is incontestable that if the reagent paper had 
 left a little moisture in the first tube, its positive reaction should 
 have been produced before the second and third trials. Further- 
 more, the presence of nitrogen oxide compounds can be disclosed 
 for a very pure specimen of tetranitrophenylmethylnitramine, pre- 
 served in a sulphuric acid desiccator at a temperature of 32. 
 
 CONCLUSIONS 
 
 a. It has been shown that if the chemically pure tetranitro com- 
 pounds 2,3,4,6- decompose more rapidly at an elevated than at 
 ordinary temperature, this may be attributed to the separation of 
 nitrous acid from nitro group 3. If tetranitroaniline be compared 
 with tetranitrophenylmethylmtramine, it is seen that the most 
 active nitro groups have also the least stability. The conception 
 of Stettbacher already mentioned, according to which the nitrous 
 acid disengaged in the decomposition of tetranitroaniline proceeds 
 from a reaction between the amido and the ortho nitro groups, is 
 inexact, for the trinitrometaphenylenediamine and the aminotri- 
 nitrophenylmethylnitramine, which ought easily to undergo this 
 species of decomposition, are very stable compounds. 
 
 Furthermore, the fact that dipicrylamine is very stable and that 
 2,4,6,2',3',4'-hexanitrodiphenylamine is not so, proves that the 
 mobile nitro group is the cause of the instability in nitro compounds 
 possessing such a mobile group. The great importance of the 
 relative mobility of this nitro group is shown by the test of 2,4,6,- 
 3 / ,4',6'-hexanitrodiphenylether, which houses the nitro group in 3, 
 and is much less mobile than the 2,4,6,2',3 / ,4'-hexanitrodiphenyl- 
 amine. 
 
 b. Tests of trinitrophenylmethymitramme and trinitrodimethyl- 
 dinitraminobenzene show that the influence on the stability of the 
 methylnitramino group, which is much less than that of a mobile 
 nitro group, is also unfavorable. When the products are sufficiently 
 pure, the introduction of these two methylnitramino groups pre-
 
 BRISANT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 25 
 
 sents no difficulty. In consideration of the tests made on the tetryl 
 it will be advisable in this preparation not to make use of too little 
 nitric acid. 
 
 c. We cannot explain completely the trinitromethylnitramino- 
 anisol and-phenetol from the point of view of absolute, as com- 
 pared with relative, stability. We believe also that a generaliza- 
 tion based on the instability of these compounds to light "is not 
 possible ; the very stable trinitromethylnitraminophenol is the more 
 sensitive to light. It is possible in these compounds that the 
 methylnitramino group becomes very mobile. However, it is not 
 possible to establish this for the basic agents employed to produce 
 the substitution of the oxyalkyl group. 
 
 One can obtain a fair idea of the value of the mobility of the 
 nitro group by a determination of the heat of combustion. This 
 determination should be made, on the one hand, for: 
 
 2,4-Dinitrophenylmethylnitramine. 
 2,4,6-Trinitrophenylmethylnitramine. 
 
 2,3,4,6 -Tetranitropheny Imethy Initr amine . 
 and on the other hand for : 
 2,4-Dinitroaniline. 
 
 , 2,4,6-Trinitroaniline. 
 
 2,3,4, 6-Tetranitroaniline . 
 
 It would suffice to give greater mobility to the fourth nitro group, 
 instead of to the nitramine in the 3 position, to augment the value 
 of the heat of combustion which will be greater for the passage of 
 the trinitramine to the tetranitronitramine than for the passage of 
 the trinitroaniline to the tetranitroaniline. The series consisting 
 of di-, tri- and tetranitrophenol is also interesting from this point 
 of view; if we note that 0.2 g. of tetranitrophenol detonates in the 
 determination of its initial explosion temperature it is evident that 
 the determination of its heat of combustion and probably also of 
 the tetranitrophenylmethylnitramine cannot be effected in the 
 explosion calorimeter. 
 
 DETERMINATION OF THE INITIAL EXPLOSION 
 TEMPERATURE 
 
 This constant of explosives is the temperature to which, through 
 a well determined elevation of temperature (see below) it is brought 
 to explode or inflame. As a means, of ascertaining the degree of 
 purity of crystallizable compounds, this constant is not of great
 
 26 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LE NNEP 
 
 value; for this purpose the fusion and decomposition points are 
 already indexes; if then determinations are made of the initial ex^ 
 plosion temperature, it is to obtain an indication of the same value ; 
 for while the decomposition and explosion temperatures are not the 
 same thing, it is not infrequently the case with the brisant nitro 
 explosives to have the same figure for the two. Only the circum- 
 stances which determine the initial explosion temperature render 
 it necessary to use a larger quantity of the substance (up to 0.5 g.) 
 to give this constant some value, because one can better observe 
 the decomposition phenomena at high temperature. 
 
 Though it has not been possible to determine if the influence of 
 stabilization is as great for brisant nitro explosives as for guncotton, l 
 it appears to us that it is still effective. Guncotton as ordinarily sta- 
 bilized has an initial explosion temperature of 180, but this may fall 
 to 120, and even lower if the guncotton is insufficiently stabilized. - 
 Trinitromethylnitraminophenol, when pure, has an initial explosion 
 temperature of 188; Van Romburgh 3 has, however, produced an 
 explosion at 120 by heating the impure compound in an oil bath. 
 But the published information on decomposition at high tem- 
 perature 4 fails to show the part played by the decomposition prod- 
 ucts which are formed. 5 
 
 As has already been said, it is indispensable in the determination 
 of the initial explosion temperature to hold to precise experimental 
 conditions in order to obtain comparable results. We have made 
 our determinations according to generally accepted methods. 6 In 
 method A, the substance is heated above 100 at the rate of 20 
 per minute, and in method B (employed in Holland for smokeless 
 powder and guncotton) the elevation of temperature above 100 is 
 at the rate of 5 per minute. 7 The temperatures found by method 
 B are evidently lower than those of method A. We, however, give 
 the preference to heating at the rate of 5 per minute, because thus 
 we can better take account of the manner in which the materials 
 behave at high temperatures. Certain substances which give a 
 
 1 Precise trials have not again been made on this subject. 
 
 2 Brunswig, "Explosivstoffe," 1909, 28. 
 
 3 Van Romburgh, Versl. Kon. Akad. Van Wetensch. Amsterdam, 23, 1314 (1915). 
 
 4 See Hoitsema, Zeitschr.f. physik. Ghent., 21, 137 (1896). 
 
 5 Brunswig, "Explosivstoffe," 29, 154 (1909); the information for guncotton will be 
 found in the bibliography. 
 
 6 Marshall, "Explosives," etc., 1915, 588; Escales, VI, 427. 
 
 7 See also Brunswig, "Explosivstoffe," 1909, 28.
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 27 
 
 more or less precise initial explosion temperature when heated at 
 the rate of 20 per minute, do not give the same phenomenon when 
 heated at the rate of 5 per minute because the decomposition is 
 more progressive. In method A we have used 0.2 g. (the quantity 
 can vary from 0.2 to 0.5 g.) and in method B we have used the pre- 
 scribed weights, that is to say, 0.1. 
 
 TABLE I 
 Table of initial explosion temperatures 1 
 
 NAME OP EXPLOSIVE 
 
 TEMPERA- 
 TURE OF 
 EXPLOSION 
 METHOD A 
 
 TEMPERA- 
 TURE OP 
 EXPLOSION 
 METHOD B 
 
 REMARKS 
 
 Trinitrotoluol 
 
 Degrees 
 321 
 
 Decrees 
 304 
 
 
 Picric acid 
 
 316 
 
 309 
 
 
 Trinitrophenylmethylnitramine 
 
 196 
 
 187 
 
 3 out of 12 with pro- 
 
 Tetranitroaniline 
 
 247 
 
 237 
 
 gressive decompn. (B) 
 2 out of 6. id. 
 
 Trinitroamidophenol 
 
 250 
 
 231 
 
 4 out of 6. id. 
 
 Trinitroamidoanisol 
 
 254 
 
 238 
 
 
 Trinitroamidophenetol 
 
 257 
 
 236 
 
 4 out of 6. id. 
 
 Trinitrometaphenylenediarnine 
 
 335 
 
 320 
 
 
 Tetranitrophenylmethylnitramine .... 
 Trinitromethylnitraminophenol 
 
 175 
 197 
 
 162 
 188 
 
 
 Trinitromethylnitraminoanisol 
 
 198 
 
 187 
 
 1 out of 6. id. 
 
 Trinitromethylnitraminophenetol. . . . 
 Aminotrinitrophenylmethylnitramine. 
 2,4,6,3 ',4 ',6 '-Hexanitrodiphenylether . 
 Hexanitrodiphenylsulphide 
 
 202 
 201 
 
 318 
 319 
 
 192 
 190 
 313 
 302 
 
 4 out of 6. id. 
 1 out of 6. id. 
 
 Hexanitrodiphenylsulphone 
 
 308 
 
 297 
 
 
 Trinitrodimethyldinitraminobenzene . 
 Dipicrylamine 
 
 214 
 258 
 
 197 
 250 
 
 3 out of 6 not observed; 
 
 2,4,6,2 ',3 ',4 '-Hexanitrodiphenylamine 
 Tetranitrophenol 
 
 287 
 251 
 
 282 
 245 
 
 the substance is pro- 
 gressively decomposed. 
 
 (B) 
 
 According to A, detona- 
 
 
 
 
 tion. 
 
 1 For more detailed information, see Van Duin, "Dissertatie," Utrecht, 1918, 89-98. 
 
 The compounds have been purified as for the tests for stability; 
 in addition, they were pulverized in advance. We have used those 
 which passed a sieve of 361 openings per sq. cm. with a diameter 
 of wire of 0.195 mm. and which have remained on a sieve of 529 
 openings with a diameter of wire of 155 mm. In this way our com- 
 pounds have been pulverized to a uniform fineness, which offers an 
 advantage in that they occupy very nearly the same space in the
 
 28 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 reaction tubes. As regards the apparatus, we have used for method 
 A apparatus of the prescribed dimensions and filled with "Woods" 
 metal only, but instead of three we have heated but two tubes at a 
 time. For method B, we have made use of the Hollandaise ap- 
 paratus which is a little larger and filled with oil. We doubt, how- 
 ever, the importance of these dimensions. We remark again that 
 the metallic bath was not agitated, while the oil bath was, and this 
 explains the more concordant results obtained under the latter 
 conditions. 
 
 Below will be found the values obtained, presented in tabular 
 form. The temperatures given are corrected, and are the mean of 
 six determinations or less. 
 
 CONCLUSIONS 
 
 a. In the determination of initial explosion temperatures it is 
 recommended that both methods A and B be used, even though one 
 apprehends that the behavior of the explosives at high temperature 
 is different for each case. It is thus, for example, that one per- 
 ceives that tetryl decomposes when heated at the rate of 5 per 
 minute, while nothing is observed when the temperature rises at the 
 rate of 20 per minute. Tetranitrophenol, treated by method B, 
 is inflamed, and ordinarily the phenomenon is vivid, while method 
 A produces generally a detonation. With the tetranitrophenyl- 
 methylnitramine, it never results in detonation but a good explosion 
 with much flame; we suppose, however, that this compound in a 
 sufficiently great quantity, will undergo detonation, 
 
 b. It is not possible to combine the stability and the initial ex- 
 plosion temperature by saying that of two explosives that having 
 the highest initial explosion temperature has the greatest stability. 1 
 For proof of this, one should compare the relative figures given for 
 the stability and those for the initial explosion temperature. 
 
 c. For explosives which decompose on fusion, the temperatures 
 of decomposition and initial explosion are generally nearly the same. 2 
 If the product does not decompose on fusion it is not possible to say 
 a priori what the initial explosion temperature is. Thus we have 
 it that tetranitrophenylmethylnitramine fuses at 147; and the 
 
 1 See Claessen, Application for patent (Holland) No. 5,240, Aug. 17, 1915. 
 
 2 It follows that the figures given for hexanitrodiphenylsulphone, viz., melting point 
 306 cor., initial explosion temperature 250-255, are erroneous (Escales, VI, 81). 
 This is probably the initial explosion temperature of the mixture of 2,4,6,3 ',4 ',6'- and 
 2,4,6,2',3',4'-hexanitrodiphenylether which is formed in the preparation.
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 29 
 
 tetranitrophenol at 140; the initial explosion temperature of the 
 latter is 250 and that of the former 175, yet these two compounds 
 are excellent examples of tetranitro derivatives. 
 
 SENSITIVENESS TO MECHANICAL EFFECTS 
 
 Of the four methods which have been employed in the test of an 
 explosive by shock, 1 notably the test by the impact machine, the 
 test by fulminate detonators, 2 the test of firing and the test of friction 
 in an unglazed porcelain mortar, 3 the first two are the more impor- 
 tant because they enable one to class the explosives more rigorously. 
 Of these two methods, that of the impact machine is the more 
 precise. This is why we have applied it to the examination of nitro 
 compounds. Will, 4 Lenze, 5 Mettegang, 6 Bichel, 7 Kast, 8 Zschokke, 9 
 have made researches having for their object the construction of 
 apparatus with which to obtain certain and precise results and also 
 to determine the conditions to which very close attention should 
 be given during the execution of the tests. This information on 
 these matters will be found in the works of Lenze cited above, and 
 in those of Kast, which give particular instructions. 
 
 Concerning apparatus and the conforming to the "surface con- 
 stant of the explosive," we refer to Kast, "Anleitung," etc., 1908, 
 1015, and Escales, "Nitrosprengstoffe," 422. The surface of the 
 stamps is, according to these authors, I 1 / 2 sq. cm., but this value 
 is not international. Thus the dimension of the apparatus of "The 
 American Bureau of Mines" is 0.785 sq. cm. 10 of surface, while that 
 proposed at the Eighth International Congress of Applied Chem- 
 istry is 1.27 sq. cm. 11 It was proposed further to make ten tests of 
 impact, taking picric acid as the standard for comparison, all of 
 these measurements to be made in the same apparatus with the 
 weight of the finely pulverized substance varying from 0.05 to 
 0.01 g. 12 
 
 1 Will, Zeitschr. f. d. ges. Schiess- und Sprengstoffw., 1, 209 (1906). 
 
 2 Bichel, "Gluckauf," 41, 1194 (1905). 
 
 3 See Kast, "Spreng- und Ziindstoffe," Braunschweig, 1909, 1017. 
 
 4 Will, loc. cit. 
 
 5 Lenze, Zeitschr. f. d. ges. Schiess- und Sprengstoffw., 1, 287 (1906). 
 
 6 Mettegang, Ibid., I, 293 (1906). 
 
 7 Bichel, Ibid., 3, 403 (1908). Marinerundschau, 16, 1345 (1905). 
 
 8 Kast, Zeitschr. f. d. ges. Schiess-und Sprengstoffw., 4, 263 (1909). 
 
 9 Zschokke, "Militarische Sprengtechnik," 1911. 
 
 10 Marshall,"Explosives, Their Manufacture, Properties, Test and History," 1915, 345 
 
 11 Zeitschr. f. d. ges. Schiess- und Sprengstoffw., 8, 370 (1913). 
 
 12 According to Kast, there is invariably employed at the "Militarversuchsamt" of 
 Berlin, 0.04 g. of the substance with a stamp surface of 0.5 sq. cm.
 
 30 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 It is also to be noted that Wohler and Matter 1 have constructed a 
 form of impact machine (Fallpendel) for testing initiating explosives 
 which is made in such a manner as to permit the compounds re- 
 sulting from the explosion to escape after detonation; this is neces- 
 sary, for otherwise the stamps are frequently broken. 
 
 On the subject of the sensibility of explosives in general, van't 
 Hoff 2 has already pointed out that definite atomic combinations 
 show always a special sensitiveness. Further, the quantity of 
 energy set free during the explosion also plays a part. For com- 
 pounds of analogous constitution, the sensitiveness appears to be 
 a measure of the augmentation of the energy set free, while strongly 
 endothermic compounds, such as mercury fulminate, lead azide, 
 etc., are as a rule very sensitive. However, there intervene also 
 unknown factors which one may expect to find among the explosives 
 which are the more sensitive of the exothermic compounds, such 
 as lead picrate. 3 Will 4 has made the most interesting research on 
 this subject on the three isomeric trinitrotoluenes. He has shown 
 that the 2,3,4- isomer is the most sensitive ; this is because it possesses 
 the most mobile nitro group in the 3 position, and consequently 
 possesses the greatest energy. 5 Then follows the 3,4,6- isomer and 
 finally the symmetric trinitrotoluene which has no mobile nitro 
 group. 
 
 In our tests, we have used the apparatus described by Kast with 
 a stamp surface of 5 sq. cm. ; we have used only a single apparatus 
 because we have had no others at our disposal. We have operated 
 at a temperature of 19 to 23 ; 6 in order to be sure of the constancy 
 in temperature of the stamps, we have proceeded as follows: Four 
 stamps were employed in a determined order and washed in acetone, 
 volatilized rapidly, and then permitted to remain during the pre- 
 vious operation. The purity of the explosives tested was the same 
 as those whose stability and initial explosion temperature were 
 measured. But as the physical state has an influence the com- 
 
 1 Wohler, Zeitscltr. f. angew. Chem., 24, 2089 (1911). 
 
 - Van't Hoff, "Vorlesung iiber theor. und physikal. Chem.," 1900, III, 95. 
 
 3 Dupre, "Mem. des Poudres et Salpetres," 11, 94 (1901). See also Brunswig, "Ex- 
 plosivstoffe," 1909, 7-14. 
 
 4 Will, Ber. d. deutsch. chem. Ges., 47, 704 (1914). 
 
 5 On this, Will has not published information; he mentions only that the three isomers 
 have almost the same heat of combustion, which is about 3660 calories. 
 
 6 Because of the bad location of our impact machine we have not operated between 
 18 and 20 (Kast, Zeitschr.f. d. ges. Sell less- und Sprengstoftw., 4, 26-3 (1909), or between 
 15 to 20 (Escales. "Nitrosprengstoffe," Leipzig, 191."), 423).
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 3 1 
 
 pounds were formed by slow and tranquil crystallization which has 
 given, in nearly all cases, hard crystals which were then pulverized. 
 Only the trinitroamidophenetol and the trinitromethylnitramino- 
 phenetol have failed to give hard crystals; these were always more 
 or less elastic. Finally, the pulverized explosives were dried during 
 24 hours over fragments of potash. We have preferred potash 
 to the usual calcium chloride because certain of these explosives 
 were recrystallized from glacial acetic acid. 
 
 For the tetranitrophenylmethylnitramine we have made a final 
 crystallization in benzene, with agitation, for the crystals obtained 
 by slow crystallization are very hard and their pulverization appears 
 too dangerous on account of the great sensibility which we attribute 
 to this body. It is true that we have pulverized this compound 
 under a glass with the aid of a horn spatula, but we have not sieved 
 it because the small particles agglutinate into boulets. 
 
 Since constancy in the thickness of the layer of explosive is of 
 great importance, we have measured the materials in a large capsule 
 containing on an average, when used for tetryl, about 42 mg. It is 
 evident that the quantities of the different materials measured in 
 this manner will not be the same. The weights are given for each 
 substance tested in the tables which follow. After having obtained 
 uniformity in the degree of fineness of the materials, we have sieved 
 them through a sieve of 1330 apertures per sq. cm., with a diameter 
 of wire of 0.105 mm. Since tetryl appears more and more to be 
 accepted as a standard, we have taken this substance as the standard 
 of comparison and we have determined the results by measurements 
 repeated each day. In the tables which follow we have recorded 
 these values beside those for the explosive tested the same day. 
 The determinations of all of these substances have been made on 
 two different days; the two values will be found in the table. 
 
 We have at first employed in our experiments a weight of 2 kg., 
 but since for about 8 explosives the maximum height of fall es- 
 tablished for this weight (60 cm. l ] is not sufficient, for these explo- 
 sives we have used a weight of 10 kg. 2 Following Kast, we have 
 determined the maximum height of fall for which there is no explo- 
 sion in an uninterrupted series of 6 trials, and the minimum height 
 of fall, for which there is always an explosion, with an uninterrupted 
 
 1 We wished to keep to the maximum height which Kast states was followed in his 
 researches. 
 
 2 We had only weights of 2 and 10 kg. at our disposal.
 
 32 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 TABLE II 
 Weight of plunger 2 kg. Temperature 19-23' 
 
 NAME OF EXPLOSIVE 
 
 MEAN 
 
 QUANTITIES 
 
 OF SUB- 
 STANCE IN 
 CAPSULE 
 IN MG. 
 
 MAX. 
 
 HT. 
 
 TO 6 
 
 CM. 
 
 MIN. 
 HT. 
 
 6 TO I 
 
 CM. 
 
 MAX. 
 HT. 
 
 TO 6 
 
 CM., 
 
 STANDARD 
 
 (TETRYL) 
 
 MIN. 
 HT -. 
 
 6 TO 6 
 
 CM. 
 
 Trinitrotoluene 39 
 
 Picric acid 36 
 
 Tetranitroaniline 39 
 
 Tetranitroamidophenol 46 
 
 Tetranitroamidoanisol 37 
 
 Tetranitroamidophenetol 37 
 
 Tetranitrometaphenylenediamine .... 33 
 
 Tetranitrophenylmethylnitramine .... 25 
 
 Trinitromethylnitraminophenol 42 
 
 Trinitromethylnitraminoanisol 39 
 
 Trinitromethylnitraminophenetol .... 34 
 
 Aminotrinitrophenylmethylnitramine. 42 
 
 2,4,6,3 ' ,4 ',6 '-Hexanitrodiphenylether. 39 
 
 Hexanitrodiphenylsulphide 47 
 
 Hexanitrodiphenylsulphone 38 
 
 Trinitrodimethyldinitraminobenzene . 40 
 
 Dipicrylamine 38 
 
 2,4,6,2',3',4'-Hexanitrodiphenylamine 41 
 
 Tetranitrophenol 40 
 
 >60 
 
 >60 
 
 >60 
 
 >60 
 
 55 
 
 54 
 
 >60 
 
 >60 
 
 >60 
 
 >60 
 
 >60 
 
 >60 
 
 >60 
 
 >60 
 
 19 
 
 19 
 
 33 
 
 30 
 
 >60 
 
 >60 
 
 >60 
 
 >60 
 
 45 
 
 43 
 
 35 
 
 32 
 
 36 
 
 39 
 
 43 
 
 43 
 
 21 
 
 26 
 
 49 
 
 49 
 
 47 
 
 44 
 
 26 
 
 28 
 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 36 > 
 36 2 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 37 3 
 36 4 
 >60 
 
 >60 
 
 >60 
 
 45 s 
 
 52 
 
 51 
 49 
 
 51 
 49 
 53 
 51 
 49 
 49 
 49 
 49 
 49 
 49 
 51 
 50 
 53 
 51 
 51 
 50 
 51 
 51 
 51 
 51 
 51 
 51 
 50 
 53 
 49 
 51 
 49 
 51 
 49 
 51 
 50 
 53 
 50 
 53 
 53 
 51 
 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >60 
 >'60 
 >60 
 >60 
 >60 
 
 1 and 2 . All six explosions complete with flames. 
 
 3 All explosions complete ; only four with flames. 
 
 4 All explosions complete with flames. 
 
 'and 6 . All explosions complete with flames, very strong detonation and great flame.
 
 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 33 
 
 series of 6 trials. 1 All these results are recorded in Tables II 
 and III. 
 
 TABLE III 
 Weight of plunger 10 kg. Temperature 19-24 
 
 NAME OP EXPLOSIVE 
 
 MEAN 
 QUANTITIES 
 OP SUB- 
 STANCE IN 
 CAPSULE 
 IN MG. 
 
 MAX. 
 
 HT. 
 TO 6 
 CM. 
 
 MIN. 
 HT. 
 6 TO 6 
 CM. 
 
 MAX. 
 HT. 
 TO 6 
 CM. 
 
 STANDARD 
 
 (TETRYL) 
 
 MIN. 
 
 HT. 
 
 6 TO 6 
 
 CM. 
 
 Trinitrotoluene 
 
 39' 
 
 24 
 
 24 
 
 14 
 
 
 24 
 
 Picric acid 
 
 36 
 
 24 
 18 
 
 24 
 24 
 
 14 
 14 
 
 
 24 
 24 
 
 Trinitroamidophenol 
 
 46 
 
 19 
 24 
 
 24 
 24 
 
 14 
 14 
 
 
 24 
 24 
 
 Trinitroamidoanisol 
 
 37 
 
 22 1 
 
 24 
 
 24 
 
 24 
 
 14 
 14 
 
 
 24 
 24 
 
 Trinitroamidophenetol 
 
 37 
 
 24 
 24 
 
 24 
 24 
 
 14 
 
 14 
 
 
 24 
 
 24 
 
 Trinitrometaphenylenediamine 
 
 33 
 
 24 
 
 24 
 
 24 
 
 24 
 
 14 
 
 14 
 
 
 24 
 24 
 
 Trinitromethylnitramlnoanisol 
 
 39 
 
 24 
 15 
 
 24 
 24 
 
 14 
 14 
 
 
 24 
 24 
 
 Trinitromethylnitraminophenetol . . . 
 
 34 
 
 16 
 19 
 16 
 
 24 
 24 
 24 
 
 14 
 14 
 14 
 
 
 24 
 24 
 24 
 
 1 At 23 and 24 cm. there was twice produced in 6 trials, a very incomplete explosion ; 
 the compounds simply inflamed. 
 
 CONCLUSIONS 
 
 a. It has been shown repeatedly that Kast's apparatus and meth- 
 ods are satisfactory for obtaining precise results. 2 
 
 b. The introduction of a methylnitramino or a fourth nitro group 
 notably augments the sensibility, while an amino group lowers it; 
 one may compare in this regard the following compounds : 
 
 Trinitrophenylmethylnitramine and trinitrodimethy Idinitramino- 
 benzene. 
 
 Trinitrophenol and trinitromethylnitraminophenol. 
 Trinitroaniline and aminotrinitrophenylmethylnitramine. 
 
 1 In the table these heights are indicated briefly as follows: Maximum height from 
 to 6. Minimum height from 6 to 6. 
 
 2 See also the detailed researches of Kast, Zeitschr.f. d. ges. Schiess- und Sprengstoffw., 
 6, 7, 21, 67 (1911). On the subjects of the picrates and cresylates and trinitro bodies, 
 it is not entirely clear why Kast tested these compounds with different heights of fall 
 when he had previously recommended their determination by "maximum height from 
 to 6 and minimum height from 6 to 6." Ibid., 4, 263 (1909). We have not encountered 
 a similar condition elsewhere.
 
 34 BRISA NT NITRO EXPLOSI VES: VAN D UIN AND VAN LENNEP 
 
 Trinitroaniline and tetranitroaniline. 
 
 Trinitrophenol and tetranitrophenol. 
 
 Trinitrophenylmethylnitramine and tetranitrophenylmethylni- 
 tramine. 
 
 Trinitrophenol and trinitroamidophenol. 
 
 Tetranitroaniline and the other tetranitro compounds. 
 
 It would be expected, a priori, that the alkylation of a phenol 
 would diminish its sensitiveness; a comparison of trinitromethyl- 
 nitraminophenol with the trinitromethylnitraminoanisol and phene- 
 tol shows clearly that this influence is great. 
 
 In addition to the nature of the group, the mobility thus intro- 
 duced into richly substituted nitro compounds has also its influ- 
 ence; 1 the mobility increases the sensitiveness. This is why the 
 aminonitrophenylmethylnitramine is more sensitive than tetryl; 
 the amino group of the first diminishes its sensitiveness. More- 
 over, the methylnitramino group is much more mobile than in the 
 tetryl and as a consequence it is possible to replace it by w-nitro- 
 aniline in the first body, while such is not the case for tetryl. 2 
 
 c. The samples of dipicrylamine and 2,4,6,2',3 / ,4'-hexanitrodi- 
 phenylamine show that the replacement of a fixed nitro group by a 
 mobile nitro group produces a remarkable effect upon the sensitive- 
 ness. 3 In the 2,4,6,2 / ,3 / ,4 / -hexamtrodiphenylamirie the nitro 
 group in the 3 position is very mobile, as shown by its direct reaction 
 with dimethylaniline. 
 
 d. The data found in the literature on the subject of certain ex- 
 plosives are not exact. It is thus that Flurscheim 4 in the following 
 table shows that tetranitroaniline is manifestly less sensitive than 
 tetryl : 
 
 Weight of plunger, 5 kg. 
 
 
 FIRST SERIES 
 
 SECOND SERIES 
 
 THIRD SERIES 
 
 Tetryl 
 
 25 cm 
 
 50 cm 
 
 35 cm 
 
 Tetranitroaniline 
 
 45 cm 
 
 W) cm 
 
 50 cm. 
 
 Picric acid . . 
 
 
 35 cm. 
 
 
 1 Van Duin, Rec. trav. chim., 38, 89 (1919). 
 
 2 Van Duin, loc. cit. 
 
 3 vSee Will, Ber. d. deutsch. chem. Ccs., 47, 704 (1914). 
 
 4 Fliirscheim, Zeitschr. f. d. ges. Schiess- und Sprengslotjw., 8, 186 (1913). On this 
 subject he has given an unacceptable explanation of these discordant values especially, 
 but little confidence can be placed in the tests of the plunger and the difference in the 
 quality of the nitro compounds. We believe that he worked each time with different 
 and impure material.
 
 BRISA NT NITRO EXPLOSI VES : VAN D UIN AND VAN LENNEP 35 
 
 Such series, for which it is evident that proper precautions have 
 not been taken, cannot lead to any conclusion. The author shows, 
 however, that the sensitiveness of tetryl and of tetranitroaniline 
 for mercury fulminate is the same and this agrees with our results. 1 
 
 Also it was shown that 2,4,6,3',4 / ,6 / -hexanitrodiphenylether is 
 quite insensititive; 2 in fact the sensitiveness of this compound is 
 greater than that of tetryl. 
 
 e. Certain of the explosives studied, the tetranitrophenylmethyl- 
 nitramine and the trinitrodimethyldinitraminobenzene ought not 
 to be included in the class of nitro explosives. 
 
 In 1906, Will 8 classified explosives according to their sensitiveness 
 to a falling weight of 2 kg. as follows, and this classification is still 
 of value: 4 
 
 1st class. Height of plunger - - 7 cm. To this class belong 
 mercury fulminate, nitroglycerine, some picrates, etc. 
 
 2d class. Height of plunger 7 -- 25 cm. Includes among others 
 the dynamites. 
 
 3d class. Height of plunger 26 100 cm. Includes among 
 others most of the aromatic nitro compounds. 
 
 4th class. Height of plunger 100 200 cm. Includes many 
 quite insensitive nitro compounds. 
 
 As we have demonstrated, sensitiveness increases by progressive 
 substitution of nitro and methylnitramino groups in the benzene 
 nucleus, and compounds formed by such substitutions belong to 
 another class than that of the ordinary nitro explosives. 
 
 Finally, on the subject of the use of modern brisant explosives 
 our previous publications should be consulted. 5 
 HEMBRUG, ZAANDAM 
 
 LABORATOIRE DE CHIMIE DE LA FABRIQUE DE MUNITIONS, 
 March, 1919. 
 
 1 See Bichel, Gliickauf, 47, 1194 (1905). 
 
 ' 2 Escales, "Nitrosprengstoffe," 82. 
 
 3 Will, Zeitschr.f. d. ges. Schiess- und Sprengstotjw., 1, 209 (1906). 
 
 4 See Marshall, "Explosives, etc.," 1915, 345. 
 
 4 Van Duin and Brackmann, Chem. Weekl, 16, 501 (1919).
 
 University of California 
 
 SOUTHERN REGIONAL LIBRARY FACILITY 
 
 405 Hilgard Avenue, Los Angeles, CA 90024-1388 
 
 Return this material to the library 
 
 from which it was borrowed. 
 
 1 
 
 RECEIVED 
 
 APR 2 4 1998 
 SEL/EMS LIBRAE
 
 Bulletin of the National Research Council 
 
 Volume 1 
 
 Number 1. The national importance of scientific and industrial re- 
 search. By George Ellery Hale and others. October, 1919. Pages 
 43. Price 50 cents. 
 
 Number 2. Research laboratories in industrial establishments of the 
 United States of America. Compiled by Alfred D. Flinn. March, 
 1920. Pages 85. Price $1.00. 
 
 Number 3. Periodical bibliographies and abstracts for the scientific 
 and technological journals of the world. Compiled by R. Cobb. 
 June, 1920. Pages 24. Price 40 cents. 
 
 Number 4. North American forest research. Compiled by the Com- 
 mittee on American Forest Research, Society of American Foresters. 
 August, 1920. Pages 146. Price $2.00. 
 
 By Edwin P. Adams. October, 
 
 Number 5. The quantum theory. 
 1920. Pages 81. Price $1.00. 
 
 Number 6. Data relating to X-ray spectra. By William Duane. 
 November, 1920. Pages 26. Price 50 cents. 
 
 Number 7. Intensity of emission of X-rays and their reflection from 
 crystals. By Bergen Davis. Problems of X-ray emission. By David 
 L.'Webster. December, 1920. Pages 47. Price 60 cents. 
 
 Number 8. Intellectual and educational status of the medical pro- 
 fession as represented in the United States Army. By Margaret V. 
 Cobb and Robert M. Yerkes. February, 1921 . Pages 76. Price $1 .00. 
 
 Volume 2 
 
 Number 9. Funds available in the United States of America for 
 the encouragement of scientific research. Compiled by Callie Hull. 
 March, 1921. Pages 81. Price $1.00. 
 
 Number 10. Report on photo-electricity including ionizing and radiating 
 potentials and related effects. By Arthur Llewelyn Hughes. April, 
 1921. Pages 87. Price $1.00.
 
 Reprint and Circular Series of the National 
 Research Council 
 
 Number 1. Report of the Patent Committee of the National Research 
 Council. Presented for the Committee by L. H. Baekeland, Acting 
 Chairman. February, 1919. Pages 24. Price 30 cents. 
 
 Number 2. Report of the Psychology Committee of the National 
 Research Council. Presented for the Committee by Robert M. 
 Yerkes, Chairman. March, 1919. Pages 51. Price 60 cents. 
 
 Number 3. Refractory materials as a field for research. By Edward 
 W. Washburn. January, 1919. Pages 24. Price 30 cents. 
 
 Number 4. Industrial research. By F. B. Jewett. 1918. Pages 
 16. Price 25 cents. 
 
 Number 5. Some problems of sidereal astronomy. By Henry N. 
 Russell. October, 1919. Pages 26. Price 30 cents. 
 
 Number 6. The development of research in the United States. By 
 James Rowland Angell. November, 1919. Pages 13. Price 25 
 cents. 
 
 Number 7. The larger opportunities for research on the relations of 
 solar and terrestrial radiation. By C. G. Abbot. February, 1920. 
 Pages 14. Price 20 cents. 
 
 Number 8. Science and the industries. By John J. Carty. February, 
 1920. Pages 16. Price 25 cents. 
 
 Number 9. A reading list on scientific and industrial research and the 
 service of the chemist to industry. By Clarence Jay West. April, 
 1920. Pages 45. Price 50 cents. 
 
 Number 10. Report on the organization of the International Astronomi- 
 cal Union. Presented for the American Section, International 
 Astronomical Union. By W. W. Campbell, Chairman, and Joel 
 Stebbins, Secretary. June, 1920. Pages 48. Price 50 cents. 
 
 Number 11. A survey of research problems in geophysics. Prepared 
 by Chairmen of Sections of the American Geophysical Union. Oc- 
 tober, 1920. Pages 57. Price 60 cents. 
 
 Number 12. Doctorates conferred in the sciences in 1920 by American 
 universities. Compiled by Callie Hull. November, 1920. Pages 9. 
 Price 20 cents. 
 
 Number 13. Research problems in colloid chemistry. By Wilder D. 
 Bancroft. April, 1921. Pages 54. Price 50 cents. 
 
 Number 14. The relation of pure science to industrial research. B T 
 John J. Carty. October, 1916. Pages 16. Price 20 cents. 
 
 Number 15. Researches on modern brisant nitro explosives. By C. F 
 van Duin and B. C. Roeters van I v ennep. Translated by Charles F 
 Munroe. February, 1920. Pages 35. Price 50 cents.