For their correspondence in the first period of separation, Van
	and Ada had invented a code which they kept perfecting during
	the next fifteen months after Van left Ardis. The entire period
	of that separation was to span almost four years ("our black
160.05	rainbow," Ada termed it), from September, 1884 to June, 1888,
	with two brief interludes of intolerable bliss (in August, 1885
	and June, 1886) and a couple of chance meetings ("through a grille
	of rain"). Codes are a bore to describe; yet a few basic
	details must be, reluctantly, given.
160.10	One-letter words remained undisguised. In any longer word
	each letter was replaced by the one succeeding it in the alphabet
	at such an ordinal point—second, third, fourth, and so forth—
	which corresponded to the number of letters in that word.
	Thus "love," a four-letter word, became "pszi" ("p" being the
160.15	fourth letter after "l" in the alphabetic series, "s" the fourth
	after "o," et cetera), whilst, say "lovely" (in which the longer
	stretch made it necessary, in two instances, to resume the alpha-
	bet after exhausting it) became "ruBkrE," where the letters
	overflowing into the new alphabetic series were capitalized: B,
160.20	for instance, standing for "v" whose substitute had to be the

[ 160 ]

	sixth letter ("lovely" consists of six letters) coming after it:
	wxyzAB, and "y" going still deeper into that next series:
	zABCDE. There is an awful moment in popular books on
	cosmic theories (that breezily begin with plain straightforward
161.05	chatty paragraphs) when there suddenly start to sprout mathe-
	matical formulas, which immediately blind one's brain. We do not
	go as far as that here. If he approaches the description of our
	lovers' code (the "our" may constitute a source of irritation in its
	own right, but never mind) with a little more attention and a little
161.10	less antipathy, the simplest-minded reader will, one trusts, under-
	stand that "overflowing" into the next ABC business.
	Unfortunately, complications arose. Ada suggested certain
	improvements, such as beginning every message in ciphered
	French, then, switching to ciphered English after the first two-
161.15	letter word, switching back to French after the first three-letter
	word, and reshuffling the shuttle with additional variations.
	Owing to these improvements the messages became even harder
	to read than to write, especially as both correspondents, in the
	exasperation of tender passion, inserted afterthoughts, deleted
161.20	phrases, rephrased insertions and reinstated deletions with mis-
	spellings and miscodings, owing as much to their struggle with
	inexpressible distress as to their overcomplicating its cryptogram.
	In the second period of separation, beginning in 1886, the
	code was radically altered. Both Van and Ada still knew by
161.25	heart the seventy-two lines of Marvell's "The Garden" and
	the forty lines of Rimbaud's "Mémoire." It was from those two
	texts that they chose the letters of the words they needed. For
	example, l2.11. l1.2.20. l2.8 meant "love," with "l" and the
	number following it denoting the line in the Marvell poem,
161.30	and the next number giving the position of the letter in that
	line, l2.11, meaning "eleventh letter in second line." I hold
	this to be pretty clear; and when, for the sake of misleading
	variety, the Rimbaud poem was used, the letter denoting the
	line would simply be capitalized. Again, this is a nuisance to

[ 161 ]

	explain, and the explanation is fun to read only for the purpose
	(thwarted, I am afraid) of looking for errors in the examples.
	Anyway, it soon proved to have defects even more serious than
	those of the first code. Security demanded they should not
162.05	possess the poems in print or script for consultation and how-
	ever marvelous their power of retention was, errors were bound
	to increase.
	They wrote to each other in the course of 1886 as often as
	before, never less than a letter per week; but, curiously enough,
162.10	in their third period of separation, from January, 1887, to
	June, 1888 (after a very long long-distance call and a very brief
	meeting), their letters grew scarcer, dwindling to a mere twenty
	in Ada's case (with only two or three in the spring of 1888)
	and about twice as many coming from Van. No passages from
162.15	the correspondence can be given here, since all the letters were
	destroyed in 1889.
	(I suggest omitting this little chapter altogether. Ada's note.)

[ 162 ]