Thank you, Dan Lehman. I was finally able to get a copy of this article, through a friend of mine, that got it from from a friend of his, that got it from you ! 
Nice find, although of a different knot !
I see that the “two way hitch” presented there - of which I was not aware of - (see attached picture) is different from all the three variations that I have posted in this thread. There is a simple reason for it, that I will try to explain. I wanted a “mid-line bend” that was able to withstand even a lengthwise pull of the attached line, so I made sure that the free end of this line remain as near the point where the ends of the nipping loop touch each other, and as near the axis of the main line, as possible ( see the pictures of this thread, and espesially the first one, where, due to symmetry, this even happens for both ends ). On the contrary, in the “two way hitch” of Wright and Magowan, the free ends leave the nipping loop as far from this point and the axis of the mainline as possible ! I am afraid that, doing this, when we pull the free end(s) from certain directions, we ran the danger to disturb and disfigure the whole knot s nub - with detrimental results in the holding power and the strength of the bend. See the attached close-up picture: If it were of the “two-way hitch”, the orange free end(s) would have been over the orange diagonal strand that also pass through the nipping loop as well, and not under it, as in the mid-line bowline bend(s).
This useful property of the “mid-line” bend is also present in the most beautiful mid-line bend by dfred. (1)
The “two way hitch” of Wright and Magowan is also implemented, in this form, in their “waist rope hitch” ( see attached picture).
Well, well : it IS a small world (of knotters, anyway) ! 
Nice find, although of a different knot !
Slightly, yes. And I should’ve referred more narrowly, to the
OP version “a” --the one symmetric one-- as like W&M’s.
W&M’s knot has the same orientations as the common knots
sheet bend and bowline when so loaded --i.e.,
of a same-side sheet bend (not “left-handed”) and common
bowline (not “cowboy”); yours is the other way 'round,
in addition to the orientation of ends vs connecting part.
To put W&M’s with a similarly “away” connecting part is
to get a much less stable knot --it shifts considerably when
loading changes! (Hmmm, I see now two versions of doing
this --ends exiting between bight legs or not.)
I find the W&M version to work okay, well enough.
–dl*
If the (orange, attached line s) point of exit is as close as possible to the “kiss” point of the (white, attaching line s) nipping loop, ( as it happens in all 3 variations of the “midline bowline” - but not in the W&M knot ), any detrimental effects of a lengthwise pull ( of the attached line ) upon the nipping loop ( of the attaching line ) are minimized. Dfred s bend, which is simpler still, also retains this useful quality. Who is going to test those 5, getting-even-more-as-the-time-passes, it seems, bends, I wonder…
dfred, can you test those two knots of yours as bases for end-of-line loops, and tell us your findings ?
The first one would be a bowline-like loop with two Myrtle collars, and the second a crossing knot loop, also with two Myrtle collars.
I think that those two loops will be as secure and strong as the common bowline, or the crossing knot loops presented at (1).
At the bowline thread (2), I came to the conclusion that we can not define and classify the end-of-line loops, if previously we have not done the same with the mid line bends that serve as “bases” of those loops. We should first examine the local properties of the nub, and only afterwards deal with the whole picture, the end-of-line loop. So, the study of your two knots takes a new twist.!
Replacing the simple nipping loop with a double, crossed-coils one - that was proved to be more stable to axial, lengthwise pull- we have a modification of dfred s midspan bend. ( See the attached pictures.)
A mid-line bend based on the Constrictor : a Strangle knot, itied on the attached line, is interliocked with a Constrictor, tied on the main line. (See the attached picture).
I have only recently realized that, if we have a complex enough, effective enough structure on the attached line, we do not need a closed nipping loop on the main line ! I was using a closed nipping structure till now, because I was driven by the similarity of a (TIB) mid line bend with the bowline - where the last thing we want, is a deformed, opened, helical nipping loop !
However, it turns out that any curve - even not so tight, closed curves as dfred and myself have been used till now - can, in fact, be utilizedo to prevent any slippage of the attached line alongside the main line.
See the attached picture, where such a wide open, helical curve is able nevertheless to hold firmly in place the more complex, tight nipping structure of the attached line. I am sure we can find even simpler midline bends than the one shown here… and the interested reader is called to try his own hand.
Read a comment on this structure on :
http://igkt.net/sm/index.php?topic=3233.msg21697#msg21697
This is an interesting thread and it highlights just how versatile the Constrictor Knot can be. The Constrictor might just be the best all around knot we have because it can serve as a bend, a loop, a binder and I even sometimes use it as a beefed up BellRinger’s (ConstrictorRinger) for my Trucker’s Hitches. The Bag Knot can do all of these things too..the fact they can be tied in the bight is what makes them so versatile.
To your orginal post Xarax, I would have to agree with Knot4U’s assertion that tying two butterfly loops together would be the best mode of operation because you want something that can take a pull in any direction. I would assume that this would be the safest method too, interesting thread regardless.
Yes, indeed… but I think that its high symmetry is also a very important factor.
You forgot to mention that the Constrictor can also serve as a fine noose-hitch, ( See the “Buntline extinguisher”, at (1) ), the adjustable bowline-like loop (2), the Constrictot-based simple lock(s) for the bowline (3,4), and the Constrictor bend(s) (5).
A possible, probably easy method of tying the Adjustable Constrictor loop.*
We first make a bowline nipping loop ( do not worry ! It is only temporary, it will disappear in due time, and will be transformed into an open helical coil…), and then we pass the working path in the wrong way … Then we form a “proper”, first bowline collar, we pass through the nipping loop again following the same path, and going over the strand it is already there, we form a second collar (around the eye leg of the standing part), and finally we pass the working end again through the nipping loop, for the last time, going over and crossing in an elbow configuration the strand of its first passage . ( This way, the strand returning from the collar around the eye leg of the standing part, will be put in an elbow, X = crossed embrace with the strand going to the collar around the standing end) . ( So, each time we move "upwards " or “downwards”, we pass the working end through the nipping loop from the same direction it had in its previous passage)( Verbal descriptions are often inadequate, my verbal descriptions are always incomprehensible, so : See the attached picture ! :)).
- I know a few only things about how knots work, and almost nothing about how minds work, so usually I am very reluctant to make any suggestions of tying methods. People should remember and tie knots according to the specific way their minds and their knots/materials combination work…I would be glad if somebody would figure out an easier way to tie this knot, that would better fit into his mind - and possibly into mine s as well !
P.S. If we want to tie the two-coils, or even the three coils version of those loops, we simply have to start from a double or triple nipping loop, and proceed as described above. I do not believe that more than one coils will be ever needed in any application, except perhaps for monofilament fishing lines.
Although one can see this as seeing a constrictor where
the nature of that binder is lost. (The clove hitch also can
lead to a nice, quickly tied (in the bight) eyeknot.)
One might prefer to use it with the other end loaded,
so, making a turn/half-hitch and then finishing away
from the hitched object (turning near, reaching far,
vs. vice versa as you show). Both look interesting.
–dl*
You mean to swap the standing end for the tail and vice versa ? I have considered this “inverted” loop, and I have decided that ;
- It is better to have the helical coil - that surrounds the Constrictor and thus multiplies its nipping force - tensioned by the standing end s strong pull, rather than by the eye leg of the bight s weaker one.
- the way the tail is exiting the nub of the inverted knot is not very satisfactory… because, in this last line of defense against slippage, this last tuck, the tail is not nipped very tightly. ( When a tail is nipped in more than one points into a knot s nub, it is better to arrange things so that it is nipped harder and harder as it proceeds closer and closer to the exit - and not the opposite. That way there would be no slack left alongside the tail s path through the knot, and the knot itself would be forced to take a more compact shape).
Here’s a boring solution: Hold the two ropes side-by-side as if the ropes are one. Tie a Butterfly loop on a bight as if the ropes are one. Tied correctly, you’ll have two loops and four ends.
Advantages:
- You already know it if you know the Butterfly Loop on a bight
- Difficult to tie wrongly
- You don’t need access to any end
- Can be pulled at all four ends, any combination of forces (according to my brief testing)
- Possible to bend FOUR separate ropes with this same knot. To understand this, imagine cutting the Butterfly loops here and then untying. Note that if you bend four ropes together the way I propose, you’ll have 8 ends and no loop.
- Not jam-prone
Disadvantages:
- Requires more rope than most other solutions here
- Not as elegant looking as most other solutions here (depending on your taste)
An understatement ! In the first place, you have tied a different knot, a TIB double loop - or a double line midline double loop. Who ordered that ? Then, economy of material is not a matter of taste, one can measure it, it is not belonging to the “de gustibus et de coloribus” subjective realm.
Now, there are dozens of midline, TIB single or double loops known, and I guess that there are dozens that are unknown. I doubt that the Alpine Butterfly would be the most economical, or the less jam-prone. We should measure the material used by all those knots, and see.
Due to their high symmetry, many of the TIB single or double loops are very interesting knots, and they can be used as bases for loops and bends, as you have noticed. However, if a knot is already bulky as an end of line loop, the same knot would appear much more bulky in the middle of a tensioned line !
Yes, indeed…It’s quite bulky, but it’s still a good default solution if the hard-to-remember preferred knot let’s you down.
By the way, if joining 4 ropes (8 ends), the Butterfly solution I described in my last post is relatively elegant.
Just a consequence of the situation (no ends).
Now, there are [i]dozens[/i] of midline, TIB single or double loops known, and I guess that there are dozens that are unknown. I doubt that the [i]Alpine Butterfly [/i] would be the most economical, or the less jam-prone.
??? How about naming some few others to compare?!
I had my doubts about Knot4U’s suggestion working for
the ends on the same side being pulled against each
other, but (also on brief checking) they seem to do so
(forming, in a sense, sorts of offset water knots (EDK)
in effect, with added bulk that seems to resist distortion).
–dl*
I agree, of course, that even a very bulky knot, can be very elegant ! Elegance has to do with symmetries and the harmonious flow of lines, not with the economy of material. However, I ,personally, would rather choose a thin/slim elegant knot, than a thick/fat elegant one !
[/quote]
I mean, the two additional loops ( slack bights ), hanging from the knot s nub. A utilitarian thing for the wet laundry, perhaps, but that was not the prescription of this thread I am afraid…
I am not going to repeat here some numbers of the ABoK s relevant long chapter on double loops you know very well ! I have a feeling that there are dozens more TIB double loops possible, because, although we can not use the two ends, we can manipulate the two bights in many ways ( so they get entangled the one to the other, around the two parts of line and possibly through some nipping loops formed in those parts ).(See one possible TIB double loop midline “base”, at the attached picture. The over-under relation of an “eye” s legs can be reversed). I was impressed that the first time I tried my hand on double loops, I met a new one ( Beginner s luck ? (1). It has happened to me to have a glimpse of that seemingly endless landscape that is waiting for us there, but I just do not have the motive to exploit it for the time being - I have other priorities. Anyway, from the practical point of view, a new mid line double loop is not such an urgently needed knot, I guess…
Let us first finish this dirty bowline job !
One might remind DL that there are pictures taken at 1826 and 1838, time flows to one direction (unfortunatelly…), and there is knot eternal life..
I think that the first Constrictor noose-hitch is more of a noose, and this second one is more of a hitch…
Hey, what the heck is the knot in a rope cargo net?