|  | 
				This set of orchestral parts, used by Mahler and others for 
				early partial and complete performances of the work, has not 
				been located and probably no longer exists. Nevertheless quite a 
				lot of its history and content can be uncovered. 
				Precisely when Strauss offered to include the first three 
				movements of the Second Symphony in one of his spring 
				Philharmonic Concerts in Berlin is not clear, not least because 
				there is a gap of five months in the sequence of Mahler's 
				letters to Strauss between August and December 1894. At this 
				time Mahler had a busy conducting schedule at the Hamburg 
				Stadttheater, was also conducting a concert series in Hamburg 
				and completing the preparation of the fair copy of the Second 
				Symphony (AF2). In his letter to Strauss of 19 July 1894 
				(GMRSB, 39–40;
				
				
				GMRSBE, 37–8) 
				Mahler announced the completion of the work (it seems unlikely 
				that the partial performance would have been discussed before 
				that date), but by 4 January 1895 he could report that the orchestral parts for the three 
				movements were finished (GMRSB, 41;
				
				
				GMRSBE, 38), and were to be used in a 
				run-through in Hamburg the following week; although it is not mentioned, the copy 
				of the score of the first three movements (ACF1) 
				must also have been ready by this date. On 27 January Mahler 
				reported that the run-through had taken place, and that the 
				material for the Symphony was ready (GMRSB, 42;
				
				
				GMRSBE, 39). At this stage it was 
				expected that the performance would take place on 18 February, 
				but Mahler had to ask for a postponement, as he was due to 
				conduct a subscription concert in Hamburg (Martner No. 59;
				BSGMOH, 331–2). 
				 
				In the absence of [CO] there is no direct way to infer which 
				score –
				AF2 or 
				CF2 (as it then was) – was the source 
				used as the copy text for the parts. However there is 
				evidence that 
				PO1 
				was engraved from [CO] 
				so a detailed collation of the first three movements in 
				AF, the 
				original layer of 
				ACF1 and 
				PO1 might offer 
				significant clues. A cursory examination suggests that (as one would suspect on practical and economic 
				grounds) once plans for the first complete performance were 
				under way, the parts copied in 1894 and subsequently revised in 
				connection with the partial premiere, were simply expanded by the 
				addition of the fourth and fifth movements to provide a complete 
				set.¹ 
				The account given by J.B. Foerster 
				of the preliminary run-through, in Hamburg in early January 
				1895, contains some fascinating details of the revision process 
				(JBFDP, 406–7; for the complete description of 
				the event, see the notes to
				
				ACF1):
					
						
							| 
				Mahler  erschien mit dem Orchestermitglied Weidich 
				[recte: Weidig], einem älteren Herrn, der die Aufgabe hatte, die von Mahler 
				während der Probe bezeichneten Abänderungen und Ergänzungen 
				vorzumerken. Unser Zuhörerkreis vernahm zuerst nur kurze 
				Bruchstücke der Musik, denn Mahler unterbracht das Spiel immer 
				wieder durch seine Bemerkungen: „Weidlich, das Violoncello 
				unisono mit dem Fagott – die Oboen streichen – Flöten verdoppeln 
				– die Harmonie in die Posaunen.” 
				Weidichs Vormerkbuch war bald vollgeschrieben. 
				Manche Abänderungen und die dynamische Details wurden gleich an 
				Ort und Stelle in die Partiturstimmen eingetragen, das Übrige 
				wurde in der Pause berichtigt. Nach dieser Vorbereitung spielte 
				das Orchester Satz um Satz ohne Unterbrechung durch und belohnte 
				zum Schluß den Komponisten mit herzlichem Applaus.  | 
							Mahler appeared with a member of the 
							orchestra, Weidich [recte: Weidig],  an old man whose 
							task it was to record the revisions and additions 
							specified by Mahler during the rehearsal. At first 
							our group of listeners heard only short fragments of 
							music, as Mahler constantly interrupted the playing 
							with his comments: Weidlich, the cello in unison 
							with the bassoon – delete the oboes – double the 
							flutes – the harmony in the trombones. 
							Weidich's notebook was soon full. 
							Many alterations and details of dynamics were 
							entered into the parts there and then, and the rest 
							were corrected during the interval. After this 
							preparation the orchestra played through movement by 
							movement without interruption and at the end awarded 
							the composer enthusiastic applause.  |  
				So both
				ACF1 and and the original three-movement 
				state of [CO] contained a layer of revisions and corrections 
				before they were used for the rehearsals in Berlin for the 
				concert in March 1895, though how accurately the score and parts 
				were collated is a matter for conjecture. A further layer of 
				revisions may have been added by Mahler during rehearsals and 
				after the March performance of the first three movements; again 
				the thoroughness of any collation of score and parts is open to 
				speculation. 
				Sometime in the spring or summer of 1895 Mahler started planning 
				the first complete performance of the work at his own expense. 
				In August, on his way back to Hamburg, he passed through Berlin and had discussions with 
				the concert agent Hermann Wolff about the project (HLG1, 327, 332), and 
				while there was told by Hermann Behn that he and Wilhelm Berkhan 
				would pay for the performance (GMLJ, 
		503;
		
				
				GMLJE, 369). The nature and extent of this support is 
				not clear, but in is his
				annotations in 
				his copy for the full score, Mengelberg noted that Behn had 
				paid for the printing costs of the parts: since Mahler's 
				other comments seem to have been concerned with preparations for 
				the first performance, one might wonder whether he was referring 
				to the copying of the orchestral parts, and/or the production of 
				chorus material (lithographed copies from writing). Whichever 
				costs this financial support defrayed, as late as 10 September 
				matters were still undecided (GMB, 
				141; 
				
				GMSL, 167), 
				but it seems very likely that despite this, the production of a complete copyist's score and performing 
				material was begun rather earlier, chorus material being a 
				priority. The last two movements were presumably added to the 
				original orchestral parts during the late summer/autumn of 1895, Between the autumn of 1895 and the appearance of the 
				printed parts (perhaps in 1898/9 or as late as 1903), the 
				resulting set of 
				manuscript parts was the only orchestral material available. 
				 
				The interrelationships between the various sources is 
				graphically summarized in a provisional
				stemma. |