in London  

Ryde, 24 September 1857  

|Ryde, 24 Sept. 1857.  

Dear Marx,  

That up there is the castle where Cromwell shut up Charles I for a time. I shall inspect it more closely on Sunday.  

Your wish concerning India crossed with the idea that occurred to me that you might like to hear my view of the business. I simultaneously found an opportunity to go over the main contents of the latest mail with the map in my hand, and voici ce qui en résulte:  

The position of the English in the district of the middle and upper Ganges is so disjointed that the only militarily correct thing would be for Havelock’s column and that of Delhi, after each has, as far as possible, picked up the detached and blockaded garrisons of the area, to unite, if possible, at Agra; apart from Agra, only to hold the neighbouring posts south of the Ganges, especially Gwalior (on account of the central Indian princes), and to have the posts lying down the Ganges—Allahabad, Benares, Dinapore—held by their previous garrisons and detachments from Calcutta; meanwhile to escort the women and non-combatants down the river, so that the troops become mobile again; and by means of mobile columns to keep the surrounding countryside in awe and to collect supplies. If Agra cannot be held, fall back on Cawnpore or Allahabad; this last point is to be held to the very utmost; it is the key to the country between the Ganges and the Jumna.  

If Agra can be held, and the Bombay army remains available, then the armies of Bombay and Madras would have to hold the peninsula proper as far as the line of latitude of Ahmadabad and Calcutta, and send out columns to establish communication northwards—the Bombay army via Indore and Gwalior to Agra, the Madras army via Saugor and Gwalior to Agra, and via Jubbulpore to Allahabad. From the Punjab, if it holds, and from Calcutta via Dinapore and Allahabad, the other lines of communication then run to Agra, so that 4 lines of communication and, excluding the Punjab, 3 lines of retreat on Calcutta, Bombay, Madras would be established. The concentration of troops on Agra from the south would thus simultaneously serve to keep the central Indian princes in check and to subdue the insurgent districts along the line of march.  

If Agra cannot be held, then the Madras army must first establish the line of communication running to Allahabad, and then advance with the troops from Allahabad on Agra, while the Bombay force advances on Gwalior.  

The Madras army seems to be recruited exclusively from the rabble and so far is safe. In Bombay they have 150 or more Hindustanis per battalion, and these are dangerous for stirring up the rest into mutiny. If the Bombay army rises, all military calculation ceases for the time being; then nothing is more certain than a colossal massacre from Kashmir to Cape Comorin. If the situation in Bombay is such that the army cannot be used against the insurgents any further, then at least the Madras columns, which have already advanced beyond Nagpore, must be reinforced and the communication with Allahabad or Benares established as rapidly as possible.  

| The absurdity of the present position of the English, brought about by the total absence of any real supreme command, is demonstrated chiefly in 2 complementary items: 1) that, split up, they allow themselves to be blockaded in a mass of small, widely scattered posts, while 2) they pin down their only mobile column before Delhi, where it not only can achieve nothing but is even going to the dogs. To order the march on Delhi is deserving of cashiering and hanging from any English general who must have known, what we have only recently learned, that the English themselves had strengthened the old fortifications to such an extent that the place can only be taken by regular siege, for which at least 15–20,000 men are required, and far more if the place is well defended. Now that they are there, they have to hold out there for political reasons; the withdrawal would be a defeat, and yet it will be difficult to avoid.  

Havelock’s troops have performed wonders. 126 miles in 8 days, together with 6–8 engagements, in that climate and season, is beyond possibility. But they are also ruined; he will probably have to let himself be blockaded in his turn after he has exhausted himself still further by excursions of shorter radius around Cawnpore, or he must return to Allahabad.  

The real line of reconquest runs up the Ganges valley; Bengal proper will be easier to hold, the race is so degenerate; the truly dangerous region begins at Dinapore. The positions Dinapore, Benares, Mirzapore, and especially Allahabad are therefore most important; from Allahabad the Doab (between the Ganges and the Jumna) and the towns on both rivers would have to be conquered first, then Oudh, then the rest. The lines from Madras and Bombay to Agra and Allahabad can only be secondary lines of operation.  

The main thing, as always, is concentration. The reinforcements that have gone up the Ganges have been totally split up; not a man has yet reached Allahabad. Perhaps inevitable in order to secure these posts, perhaps not. In any case the number of posts to be held is to be reduced to the minimum, the forces concentrated for the field. If C. Campbell, of whom we so far only know that he is brave, wants to distinguish himself as a general, he must create a mobile army coûte que coûte, whether Delhi is abandoned or not. || And no situation is so desperate, where there are in summa summarum 25–30,000 European soldiers, that at least 5,000 could not be scraped together for a campaign, replacing their losses from the garrison troops picked up from the posts. Only then can Campbell see how he stands and what kind of enemy he actually has before him. The odds are, however, that he will go and squat before Delhi like a fool and see his men go to the dogs at the rate of 100 a day, whereupon it will again be all the ‘braver’ to stay put all the more, until they are happily all gone to the devil. Brave stupidity remains the order of the day.  

Concentration of forces for the field in the north, vigorous support from Madras and if possible from Bombay—that is everything. Even if the Maratha princes on the Nerbudda fall away, that does not matter much except as an example; their troops are already with the insurgents. The utmost, however, that can be achieved is to hold out until the first reinforcements arrive from Europe at the end of October. But if a couple more Bombay regiments rebel, then tactics and strategy are at an end; there lies the decision.  

I am going to Brighton on Tuesday at the latest, and from there on Wednesday evening at 10 o’clock to Jersey, but I will write to you again more precisely, and hope you will come. Tomorrow I’m off to The Battery etc. Today I drove round the island, and as I last night again worked hard until 3 o’clock, I mean to sleep it off today.  

Yours