Constrictor bowline(s)

xarax · 18 June 2011 01:03

Mark Gommers, in his fine article "An analysis of bowlines" (1), presents a bowline he calls "the Constrictor bowline" (See the first attached picture, fig.16 of the article. Note : The corresponding "rear view" is missing. In its place, it was inserted, by a typo mistake, the rear view of the "reversed" Clove hitch bowline, i.e. fig.13 )

In a Constrictor bowline, the nipping loop of the simple bowline has been replaced by a Constrictor hitch. However, at this article, this Constrictor hitch is oriented in a “reversed” way. (See the second attached picture) If we use a non-reversed Constrictor hitch, we have the Constrictor bowline shown here ( See the third attached picture).
In (2), I have shown a picture of the “Buntline extinguisher” :), a simple noose that is tied as quickly and easily as the Buntline hitch. It was pointed to me, by all the participants in the discussion there, that this hitch is, in fact, a type of Constrictor hitch around the standing end of the noose, and not a genuine fig. 8 based hitch. Trying to figure out how we can name this hitch, I remembered the “Constrictor bowline” shown in Mark Gommers article, and I realized that the knot presented there is, in fact , the “reversed” Constrictor bowline, and not the “normal”(?) Constrictor bowline shown here.

Inkanyezi · 20 June 2011 08:08

I think it is confusing, and counterproductive to apply the name of the bowline to just any tangle that ends with a loop.

Nomenclature of course is built on convention, and knotting matters are rather flimsy when it comes to descriptions. Therefore, in the paper that is presented, by Mark Grommers, knots are laid out for photogenic appearance, disregarding distinctive features and their true nature.

The bowline, the simple one that has been used for centuries, has four basic forms, all incorporating a single distinctive feature, that was usually not described in literature or when teaching it, the turNip. The turNip is a single round turn, that upon load orients itself in a spiral almost in line with the standing part. What holds the knot together is the force that opposes the straightening out of this spiral. This is achieved in the bowline by the bight collar around one of the parts, either the standing part or one of its legs, as well as the opposing force from one of the loop legs.

By adding complexity to the distinctive feature of the bowline, other knots emerge. They may be simple extensions, as when you make another round turn or adding another “twist”, which will effectively form something resembling a clove hitch with the turns separated, but all extensions that eliminate the turNip and force the knot into another form, will not to my eye constitute bowlines. Thus the “clove hitch” bowline is still a bowline, provided the twists are sufficiently separated to permit the one closest to the standing part to form a turNip. The two round turn knot also in some way preserves the turNip function, but further extensions as “reversed clove hitch” or constrictor deviate too much from the original pattern to share the name in my opinion.

Part of the confusion comes from the photogenic appearance, that does not show the nature of the knot, but only a geometric pattern to explain how it may be contrived by clearly indicating the orientation of its parts at one stage of its making.

One might be more exigent regarding the name, not accepting any change of the knot pattern to share the same name, but as there are eskimo and cowboy bowlines, which share the turNip feature, I am prone to accept any loop which has a turNip that is locked by a bight collar around one of the parts as a bowline. That gives four basic bowlines, the right-hand bowline, the cowboy bowline, and those two forms in eskimo style.

There are several ways of tucking the extra Janus collar, which introduces some extra material into the turNip and adds friction to restrain the end from slipping out of the knot. The Janus collar is the most sensible addition to the pattern, as it extends from the loose end and preserves the turNip in its original orientation. Yosemite finishes, to me look mainly cosmetic. the “fig 8” that is formed is not loaded on the end, and does not pass through any nip. It is more likely just a different orientation of the end, taking a half turn around the cook’s leg and tucking it through a rather loose bight. The knot is still a bowline, but the extra complication is not secured, so it adds little to the knot besides appearance.

When extending the knot pattern, there is also a risk of removing important features from the original knot. One feature that has made the bowline particularly popular is the ease of untying. Quite naturally, any extension may make it more difficult to untie or even prone to jam. particularly a reversed clove hitch or constrictor might be difficult to untie after severe loading, thus eliminating one of the basic features of the original bowline apart from the evident lack of a turNip in their structure.

The most sensible extensions would preserve the turNip, resist straightening of its sprial form, and secure the end as well. The Yosemite finish preserves the turNip, but does not secure the end or add resistance to straightening the spiral form of the turNip, because the end is not nipped. The Janus finishes however puts the end another time into the nip, thus securing the end, while still permitting the turNip to form naturally from the standing part. The advantages of the Janus finish are that it indeed does add security and that it does not compromise the original knot, neither in its simple forming by a twist of the hand, the natural forming of a turNip upon loading, nor its ease of untying by “breaking the neck”, a twofold process in the Janus variant.

Grommers does not show any eskimo bowline, which is the form most resistant to ring loading. My option for a more secure bowline would be the eskimo with a Janus finish. The knot has one single turNip, is resistant to ring loading, has extra material that widens the initial turn from the standing part, and it secures the end by passing it through the nip after making a second collar around the standing part. It is easily tied, easily inspected and easy to untie even after severe load. It is not prone to capsize like the original bowline. Further testing is needed before recommending it for SAR work.

xarax · 20 June 2011 11:09

Thank you Inkanyezi,

You are mistaken here. It is not “just any tangle” it is shown in this paper ! They are end of the line loops that have the three specific characteristics of the bowline family of end-of-the-line-loops : (Read also (1))

[i]"There are three, and only three elements that characterize a bowline, in relation to any other end of line loop:

The knot tied on the standing part s leg, should be a slip knot. Any sailor will laugh with an end of line loop that is not completely untied like the bowline.
This slip knot should include one, at least, nipping loop, which secures the tail.
The tail should form one, at least, collar. " [/i]

In this sense, which is the least common denominator of all bowlines, the end-of-the- line-loops presented by ark Gommers are, clearly, forms of bowlines. Certainly, even if you do not include them in a more narrow definition of a bowline, they are nevertheless not accurately described by this “just any tangle”…
Having said that, I think that the rest of your article here is very well written. I believe we should give more attention to the structural elements of the knots, and less to their names and classification. The “Constrictor bowlines” , the “reversed” one, presented by Mark Gommers, and the one presented in this thread, deserve to be examined as end-of-the-line loops tied on a slip knot - that incorporates a nipping loop - on the standing part, and having one collar. If they deserve being called “bowlines” or not, is only of a secondary importance. According to my definition given above, they are. To a more narrow definition given by you, they are not. Let us compare them on knotting terms, as secure. easy tied and untied end of the line loops, and not on semantics terms.

http://igkt.net/sm/index.php?topic=2897.msg17389#msg17389

Dan_Lehman · 21 June 2011 05:58

I don’t understand what is meant by tying a slip-knot
on the “leg” of the SPart!? (A slip knot is an overhand
finished with a tucked bight vice the single-strand tail.)

–dl*

ps: A sailor will not be laughing if he has to deal with firm
kernmantle rope and his own behind is at stake – he might
want to quick learn a thing or two more about the bowline !

Dan_Lehman · 21 June 2011 06:34

As was pointed out by Xarax, it isn’t just “any tangle that ends
with [an eye]” --but, rather, one formed in such a way as to
have some core knot through which a collaring bight is
reeved. In this sense, the nomenclature is rather intuitive
and instructive (in the cases of “clove”/“cow”/“constrictor”/…
one might have to further figure out how that “core knot”
is to be oriented).

BUT, I too have tried to form a more restrictive definition
of a “bowline” --in part, to respond to such things as
the How many bowlines are there? researches (such as
Xarax links to above with his “(1)”). One should know what
it is one is setting out to count, after all! And, yes, among
those lists typically come knots such as you disclaim; we
have also seen “new” knots given the moniker, as though
to try to elevate their worth immediately --marketing stuff.

My thinking has been that a bowline has as its fundamental
component a turNip --which might be a round turn (or more),
from which the rope flows into the knot’s eye. I then distinguish
“anti-bowlines” from “bowlines” based upon which direction the
tail goes through this central nip. I’m willing to give up the collar,
and so include the so-called “Myrtle” eyeknot. And I’m not keen
to insist on a spiral vs. less-inclined loop, even a spiral
oriented in the opening way (i.e., wanting to straighten),
which I have seen bowlines of the old sort capsize into,
anyway (though at that point, I’m going to call them former
bowlines ).

I’ve also wanted to distinguish between such bowlinesque knots
that lead from this nipping part directly into the eye and those
that do something else, such as … well, make a 2nd turNip as
for the clove-hitch version of the water bowline or some
other knots I’ve discovered in which the rope here leads to a
collaring of the eye legs --a feature that might make the difference
between holding and slipping out, in slippery HMPE rope.
But this gets complicated. “false bowline” has been one name
that has occurred to me, but I think that’s overly strong in
connotation; rather, just some sense of delay/intermediation
is wanted for his indirect path to the eye.

The Eskimo bowline poses a problem: the closing bight
which collars an eye leg can be drawn snug such that the
core knot is a crossing knot (roughly a Munter hitch ),
and that seems reasonably distinct from the turNip, as it
sort of collars itself. --and this is a matter of collar size,
nothing more!

Yosemite finishes, to me look mainly cosmetic.

I’ve never been a fan of this, either, which seemed to ask
too much of a stiff rope in taking a 1-diameter wrap with
the tail, and then only wedging it into the not so terribly
nipping-gripping collar. But it has a long history with rockclimbers
and must work adequately well; only so much extra securing
is needed in order to arrest the loosening to which the common
bowline is prone. One fellow --viz., Craig Connally, author
of [u]The Mountaineering Handbook– goes so far as to claim
that his testing shows the YoBowl stronger than a fig.8
eyeknot --I call “bogus”!

the "fig 8" that is formed is not loaded on the end, ...

Actually, there was just some chatter on an ACA canyoneering forum
about tying with the bight so that this “end” is a bight-end and it
can be loaded, as an adjacent tie-in point to the SPart.
(Rob Chisnall showed doing this as an aid to providing climbing
instruction, long ago.)

The most sensible extensions would preserve the turNip, resist straightening [u]of its sprial form[/u], ...

Hmmm, you know, I think that “spiral” isn’t right, here :
rather, the would-be spiral is twisted against the direction
of opening (initially, from a mere turn start, more obviously
being a spiral than a(n imperfect) ring)) in the common bowline,
and my “anti-bowlines” are challenged to come close to this
sort of orientation, usually.

Grommers does not show any [i]eskimo bowline,[/i] which is the form most resistant to ring loading. My option for a more secure [i]bowline[/i] would be the eskimo with a Janus finish. ...

I’m not sure what you mean by the finish, but, yes, tying what
Grommers shows in #32-33 but such that the tail closes the eye
by flowing through the turNip in the opposite direction
–i.e., connect the tail to his shown end, and leave as the new
end the old entry strand-- seems a good knot. In a way, it lets
one be “a winner either way,” no matter which side of the turNip
you enter, just be sure to collar both a leg & the SPart (realizing
that this “collaring” requires that one go around the appropriate
part given the particular direction of entry!).

As for resisting ring-loading, the “left-handed” bowline does that
even without a further tuck.

–dl*

Inkanyezi · 21 June 2011 07:23

I am aware that the left-handed bowline is resistant to ring loading, but even though it is, altarnate ring and SP loading may loosen it a bit at a time, whereas the Janus finishing of any bowline will make such creeping action less probable. The drawback that may be obnoxious to a sailor is that a Janus eskimo bowline is more difficult to untie; not really jamming, but very firm when it has been loaded. That feature is not so evident in kernmantle, and a climber might prefer a knot that becomes more hard to untie rather than one that slips.

Transminator · 21 June 2011 09:02

On what grounds? Do you have conflicting test results?
An expert such as yourself does not dismiss other people’s findings just on the basis of bias (because you are not a fan of the yosemite finish and thus don’t WANT it to perform well) I hope?!

I think the secret of the yosemite finish is in the dressing. When you dress and tighten it propperly before use the nip and grip thus created is more then enough to secure the end. When we look at it closely, we see that this end forms are collar underneath the turnip. When pulling on the end and the loop, this collar and the bight through which the end is tucked, tighten. The collar tends to push the turnip upwards, which squeezes the bight together. Also when dressing/tightening the knot before use, you should pull the “double-collar” (formed by the turnip and the wrapped around end underneath) upwards to further tighten the bight. That way you create a considerable amount of grip (bight) and friction (wrap around the leg). Enough to secure the end for sure.
That this nip and grip is brought about manually in the dressing process is irrelevant. The end result is what counts.
Yes, the mechanics of the knot do not further clamp down on the end when loaded, but as you pointed out “only so much extra securing
is needed in order to arrest the loosening to which the common
bowline is prone”.

I think the above is why it is performing as well as it does. It is not just cosmetic. It might appear that way as the nip and grip is applied manually in the dressing process and loading the knot does not further improve the finish by applying addition nipping force but I wonder whether this actually can be an advantage here. If a lot of force is applied to the rope by a nipping loop, can this not also be the reason why the rope breaks at this point because the material is weakend?

According to the german wikipedia article the yosemite bowline (called Bulin 1.5 in german speaking regions) has the highest breaking strenght of all loop knots used in climbing. According to the article the rope’s retained tear strength is 63% for the Figure 8 and 67% for the yosemite bowline. The tests were carried out by the DAV (Deutscher Alpenverein = German Alpine Association) on dynamic climbing rope and the Company Edelrid on static rope. Interestingly there is not much difference for the bowline, yosemite bowline and the figure of eight on static rope. (See attachment below)

I have tied the “janus eskimo bowline”, after reading your comment, by wrapping the end around the left leg of the loop before tucking it up through the turnip. I asume that is the way you inteded it, is it not?
I am not particularly impressed by the way it dresses and it does not seem to be a good knot for “normal” loading of the loop. In most cases, the load of the loop is probably not ring loading, therefore the focus should be a secure bowline for normal loading (legs approx. parallel) + ring loading and not the other way round. The normal double-bight (Janus) bowline does exactly that. You also have the added advantage of dealing with a “normal” bowline with all the different tying methods available.

The double-bight (Janus) bowline has all the features one could ask for in a bowline. It is simple to tie, easy to untie (the yosemite is not actually as easily untied as some sources claim. The double rope diameter through the tightened bight (collar) is counter-productive for that purposes) and is secure (also for ring loading).
Also the test results indicate that the Janus holds what it promises, i.e. to be a more secure bowline.
One final thought: The double-knotted (aka round turn bowline) does, perhaps, not get enough credit as it deserves. It also is very secure for normal and ring loading, works for both the left and right handed version and is still easily untied, at least in my experience. I frequently found the yosemite bowline more difficult to untie then the double-knotted bowline.

xarax · 21 June 2011 09:26

Perhaps I used a misleading expression…A knot which, when pulled by its ends, slips until it is untied completely, and becomes an unknot.

Kernmantle ropes are not used as mooring lines so often…

xarax · 21 June 2011 10:36

I believe that the collar is an essential element of the bowline family of the end-of-the-line loops. The “Myrtle” loop is not quite a bowline, it is something less of a bowline, and/or it is an ill-tied bowline…It would be interesting to measure the holding power of those two loops, and compare them directly. I guess that the simple “Myrtle” will hold around half of the load that the simple common bowline can hold, which would justify my definition. ( Better define the bowline family of knots in a way that would include more numbers of knots we would have wished, than knots with less qualities we should have accepted.
I enjoy the discussion here about bowlines in general, but I would also like to listen opinions about the particular “Constrictor bowline(s)” shown in the original post of this thread. Has anybody any experience with any of those two variations ?

SS369 · 21 June 2011 13:30

Hi xarax,

the other day when climbing I tied/adhoc tested this constrictor bowline. It did not come untied, but it raises the thought(s) as to why I would need this configuration. I personally have never had a bowline of any kind come untied, but I also have never used it in a situation that would bring the rope to that severe loading. It is not my “go to” knot for most tasks.

Back to the “test”. Over a branch and relatively close to the ground I used it for harness tie in and then loaded it with my weight. The rope was of dynamic kermantle construction and so it was that I “bungeed” a few times.
Dressed and tightened prior I found the affair so much more aggravating to untie than it’s simpler predecessor/parent knot.

For a ring loaded use I would choose something more fitting with that set of circumstances.

So this leads to the question: Is it security or strength that this tangle is supposed to improve?
I have no test basis for the assumption that the rope will break similarly in the same local, but, that is the conclusion I come to.

You think so?

SS

xarax · 21 June 2011 15:06

Thank you SS369,

I have, in severe loading AND conditions (water, oil, dynamical loading due to wave action, etc..) Never ever did the simple/common bowlines give an inch. However, it was always with braided marine ropes, not kernmantel climbing ropes.

I have no idea. I just saw the “reversed” Constrictor bowline in Marc Gommers article, with the mistaken rear view, and I thought it is better to report that there is another variation of the same knot. I am no expert in bowlines. I think that many people in this forum have a more extended knowledge and experience of this field, and I am waiting to learn more from them.

No, not necessarily, as I have said also in the case of the “Constrictor bend” (1). The Constrictor-based mechanism ( even in the case of the loop, with one and one knot only ) is a very complex thing. It behaves very differently from the single or double nipping loop, the tension and the compression forces along the rope strand are distributed very differently, so I will really be surprised if it brakes in the same area.

Why do we need this bowline ? For the same reason we need all the others ! ( Mark Gommers has a lot, but there are still many others… I have published 4 more in this forum (2) ) To me, it is interesting to learn and test all the simple knots / inanimate rope machines that could possibly exist, and might have a practical use by intelligent living machines, in this planet, or in any else !

SS369 · 21 June 2011 15:48

Glad you didn’t get tossed from the boat xarax.
With all those contaminants I can see failure looming anyway!

I am all into the explorations for sure. I think the knot’s security may have bee raised for given circumstances, possibly sacrificing untie-ability.

As for breaking, I go by what I have read that has come across my eyes in many searches over the years and though the loop sometimes breaks before the SP does, it is the higher average that is more commonly reported that the entry point is the referred place of rupture, as in other knots as well.
First curve intersection forces sets the tone.

Of course the testing is mostly straight line and no other variables added, except in the “home grown” types of rope destruction.

Same first curve forces in the presently offered versions don’t you think?

SS

xarax · 21 June 2011 18:01

Yes, evidently, but…
I try to imagine being a molecule bonded through intra-molecular forces to my neighbours into the macromolecule chain. I am squeezed by the one side and pulled by the other, heat, in the form of molecule vibrations and ultrared photons is coming and going, somewhere accumulating and elsewhere dissipating, impurities make some fibres weaker and some stronger, the kern and the mantel are twisting and moving to opposite directions…
So, me, the poor molecule, I am totally confused…How the hell am I going to predict where I am supposed to settle ? The more complex the form of the knot, the less certain are our predictions. If we could simulate the rope and the knot, and see the flow of the forces and the heat inside it…But why bother ? We can test the 10, or 100 bends/loops/hitches 10 or 100 times with 10 or 100 different materials…A properly equipped laboratory would perform all those tests in one to ten years. So what ? I guess that tests of weapons, that kill people, are 10 or 100 times more expensive than would be the tests of ropes and knots, that save people !

Dan_Lehman · 21 June 2011 18:20

Well, one can hope.

On the basis that Connally himself has no clearly presented
data to support it. And he made some dubious statement about
pure pull-test results (i.e., where one would determine strength
of each individual knot and then compare those) being the
“wrong kind” of testing, in preference to A-vs-B testing (huh?).
AND in that values of various test results are typically better
for the fig.8 AND good (i.e., hard to beat). And test
data can be all over the place.

Beyond that, I cannot see what would make a difference
between the common & Yosemite bowlines. These
have a turn around two rope diameters, and little friction
on the SPart mitigating that. Whereas the fig.8 eyeknot
has the twin-parts rub of its collar to help take load off
of the SPart. At best, I can see that in some dressing of
the Yosemite bowlne the tail --which could be said
to move in relation to the tali-leg on the draw of the SPart
like one sitting and crossing legs (the tail brought up over
the eyeleg)-- will be held from further movement by the
SPart, and thereby go a slight way towards absorbing
some of the load (in its relative compressive softness)a
and easing the high friction point of the SPart’s turn
around the tail-side eyeleg. --but that is a SLIGHT change.
(And depends on some particular dressing in accommodating
cordage.)

When you dress and tighten it properly ...

Where is this proper dressing presented?

I don’t follow your details about dressing, and especially about
another “collar” --don’t see it that way, as the tail turns around
the tail-side eyeleg and crosses back OUTSIDE of the turNip
to exit through the only “collar” to my mind.
And you end this prg. talking about added security, but you start
with addressing a benefit to strength --and in rockclimbing cordage
the only security at issue is slack-security (so it’s not like some
knots that might slip and so cheat strength testing of higher values).

According to the german wikipedia article the [i]yosemite bowline[/i] (called [i]Bulin 1.5 [/i]in german-speaking regions) has the highest breaking strenght of all loop knots used in climbing. According to the article the rope's retained tear strength is 63% for the [i]Figure 8 [/i]and 67% for the [i]yosemite bowline. [/i]The tests were carried out by the DAV (Deutscher Alpenverein = German Alpine Association) on dynamic climbing rope and the Company Edelrid on static rope. Interestingly there is not much difference for the [i]bowline, yosemite bowline[/i] and the f[i]igure of eight[/i] on static rope. (See attachment below)

Thanks for that find.
The images are tiny, and one can wonder (regardless --who says
the image in a report has any resemblance to what testers used?!)
at what geometries actually met the test device. AND how many
tests were run, what the statistical significance was, and so on.
And this is apparently in one particular dynamic, “static” rope, respectively.
One can find reports (Lyon for HSE) where several ropes were used
and different results obtained.

Let me re-state : imagine what Grommer’s shows at #32-3 [newly 30-1];
now cut the tail-side of the eye a little away from its (re-)entry
into the knot --this becomes your new “end”-- ; connect your
cut eye part to the end in the image --this is what I meant
about going in the opposite direction. THAT seems a decent
knot, dressed/oriented just so (i.e., vis-a-vis the crossings through
the turNip, which could occur in various ways.

–dl*

agent_smith · 23 June 2011 02:48

I have updated my article on Bowlines (go here to download http://www.paci.com.au/downloads_public/knots/02_Bowlines.pdf )

Since Jan 2009, i have extensively used the EBSB variant of the Bowline in both climbing and rescue applications.

The knot is both secure and stable. I maintain that strength alone should not be the main factor in deciding upon a suitable knot for life support applications. In my view, the properties of security and stability are of greater importance.

I have removed the image of the constrictor bowline from my work…will try to obtain new and correct image some time soon.

I acknowledge that my EBSB is a derivative of Dan Lehman’s EBDB Bowline - I sought to capture the tail by tucking underneath a loop and collar which I have found to be very secure. There are also 3 rope diameters within the nipping loop - which in theory should increase the radius of the nipping loop (need a better identifier please) thereby reducing its stress and strain (my theory). I remember that Dan Lehman was not a fan of my EBSB variant - but never fully understood why?

I also wanted the tail to exit along the same trajectory as the standing part - which would allow some user groups to add a ‘stopper knot’ if they wish - eg double overhand.

I stopped testing on the basis of pure strength due to time and cost factors…but also because I believed more in knot security and stability. Still weighing in on whether to continue on strength testing…

Mark Gommers

Dan_Lehman · 23 June 2011 05:27

Thank you, it came without problems we encountered previously
(something about nuances with PDF).

Since Jan 2009, i have extensively used the [i]EBSB[/i] variant of the [i]Bowline[/i] in both climbing and rescue applications.
The knot is both secure and stable. …
I remember that Dan Lehman was not a fan of my EBSB variant - but never fully understood why?

Not sure that I fully understood this unillustrated (then) version,
but of the non-Yosemite’d knot, I don’t like the EBSB because
the end-binding turns around just 2 diameters and thus has
less security; for the EBDB, it turns around three diameters and
holds well (except in some firm & slick PP laid cord someone sent
to me on account of how intractable it was (“Devil’s material!”,
he cursed at it!)). Now, in the Yo’d version of the EBSB, yes,
there is a 3rd diameter; but the working of the knot is a PITA
–one cannot simply pull on the tail to tighten its binding,
for that bloody tail is further turned and tucked beneath
the very “end-binding” in need of tightening to bind!
(Rather, one much fiddle with the collar, and thereby
either have a bigger, too-loose collar or feed material
back into the eye.)

You are clearly in luv with the bloody “Yosemite” finish,
which I have never liked, for its hard turn around an eyeleg,
and less than seemingly decent security after all that. (And
some vulnerability to the wrap going into unwanted position
within the nipping loop.)

I also wanted the tail to exit along the same trajectory as the standing part -- which would allow some user groups to add a 'stopper knot' if they wish - eg [i]double overhand[/i].

Holy Overload, Bat Man!!!
The “end-binding” is supposed to be an alternative to back-up
knotting, not another nail in the coffin! (Btw, prefer “strangle” to “d.o.”.)

Btw, newly realized (at some point re your writing) information is that
the “Janus bowline” was first (so-far known) presented by the
Wright & Magowan 1928 Alpine Journal article in which they also
presented the Butterfly, Sennit, & Reever . Their version differs
only slightly, in the crossing of the tail with itself --reversed over/under.
So, you might note that as a bit of history to append to Prohaska’s
separate later discovery.

–dl*

xarax · 16 July 2011 08:47

A Janus bowline version can, in fact, differ a lot ! We can have a left- or a right-handed bowline (=2), the tail can pass - when it goes through the nipping loop for the second time - over or under itself (=4), the second collar can be a left- or a right-handed one (=8), and, finally, the tail can pass - when it goes through the nipping loop for the third time - in between 4 different openings, left there from its two previous passages (=32)… to my latest counting. ( And I am not considering here the very similar knots, where the tail, when it goes through the nipping loop for the second time, can transverse it through the opposite direction…) So, 32 slightly different, but different nevertheless Janus bowline versions, are patiently waiting for the brave knot tester, to tell which is more stable, secure and strong !