The Sketch of the Analytical Enigine.. uses some typgraphical conventions not common today, and there are a few apparent typographical errors.
It is available as a facsimile unter archive.org (processed version removing the background: menabrea_lovelace1843_analyticalengine.pdf) and has been typeset anew in several places, e.g. in Charles Babbage and his CALCULATING ENGINES, edited by Philip and Emily Morrison, Dover Publications 1961, The Works of Charles Babbage, edited by Martin Campbell-Kelly, Vol. 3, London 1989, and online at http://fourmilab.ch/babbage/sketch.html. Follows a list of typographic errors which are not generally all corrected (pagenumbers refer to the original publication):
The well-discussed case of the cos on p. 637, where the phrase when the cos of n=∝ has been foreseen … should clearly read the case of …
In the first equations of Note E (p.712), most reprints correctly use the upper index '1' instead of the prime as in the original; the equations referred to are not on p.679, but on p. 684
Also in Note E p.714, equation 3, there is missing a cos in the second term of the right hand side; the formula should read (see page 723):
`cos n theta * cos theta = 1/2 cos [(n+1) theta] + 1/2 cos [(n-1) theta]`
In the final table ending Note G, in operation 4, the operands have to be exchanged: V4 has to be divided by V5
In the same table in line 21, it must read 2V12 instead of 0V12 in the result column
In the same line, the zero in column V11 must be dropped. If the loop is repeated, the value is needed in the next round.
In the same line 21, in the 5th column (Indication…) the second line 0V12 = 2V12 is exceptional, as elsewhere in this column only Variables acted upon from the 3rd column are used; systematically, it would be `0V11 = 5V11.
In operation 24, the negative value of V13 has to be transferred to V24, i.e. V24-V13 must be calculated, and in the 6th column read =-B7
Also in the same line, in the (forth to last) column for V13 the digit zero must printed, as is given in the 5th column, according to the rule that intermediate variables have to be to set to zero finally.
The last signature (p.731) is not A.A.L., but A.L.L., which is most probably only a typeset error.
Note that in line 25 in the result column and in the next one right to the equals sign should read 2V3 instead of 1V3; however AAL argued that the upper index is correct, as this value is for the next round of calculations, and thus must have the upper index 1 for a starting value.
Note also that instead of nested parenthesis, parts of equations are overbarred, e.g. (4.) on page 716.
Martin Campbell-Kelley's edition corrects all, but not the wrong division in operation 4 of the table for the calculation of Bernoulli's numbers, although having used a simulation as well as Petrenko, refering to Tee's review, as of the footnote on p.159. Probably both corrected the error during testing, and missed a note on this.