A knot by Design

[DerekSmith]

I have been interested in the numerous friction hitches we have at our disposal.  Claims are made for each including the amazing claim made for the Icicle Hitch that it can hold fast to a tapered spar.

Examination of the various hitches showed that they tended to have similar conformations.  They would have a load application section, a set of coils for grip and an anchor.  Generally the anchor is formed by taking a line from the anchor, down, across the coil to clamp the the anchor and the coils in place, to the load tension point.

Generally the limitations of these various knots stemmed essentially from the clamping line pulling the whole knot forward and failing it.

The knot presented here was designed purely from principle and to this end, if the knot has not already been documented and classified, then I would like to name it the "KC Hitch" - in honour of the person who tries relentlessly to educate us all into the world of knot leverage.  This knot is based on two principles. The first is the Power of Wrapped friction law perfectly explained by Roo_Two on his web page http://www.geocities.com/roo_two/friction.html  .  This law describes how a small holding force is increased exponentially with the angle of contact - roughly to the power of three times the coefficient of friction for every complete turn.

The second principle is that of Angled Leverage - continuously promoted by KC.  This principle shows us that small tangential movements of a line away from the shortest path have a huge mechanical advantage approaching infinity as the line is first starting to be moved away from the shortest path.  Most of us will have utilised this principle, tightening a load by pulling sideways on a rope or by standing on a towline to budge an obstinate load.

Both of these principles are at work in all of the documented friction hitches, however, in designing this knot the principles were used to create the knot form rather than happen by accident by trial and error tying of cord.  As a consequence, this friction hitch should be approaching the optimum for friction hitch abilities.  However, only stringent Peer Review and in field use will show if theory alone has given us a better friction hitch.

The Design Stage.

The Wrapped Friction Law can only amplify a gripping force, so any use of this law must start with a finite gripping point, so the knot must have three sections or functions:-

Part 1 : The anchor - this will provide the embryonic gripping force which will be amplified by the coils.
Part 2 : The gripping engine - this will utilise coils to exponentially amplify the anchor grip to the level needed to hold the object and transfer all the working force into the held object.
Part 3 : The force engine - this will allow the applied force to be transmitted into the knot, to tension it and to allow for surface geometry and movement of the knot.

To start then, I used the best gripping anchor I know - the Constrictor.  I tied a constrictor onto a length of dowel and yes it was able to generate a small gripping force which could be fed into an amplification coil.  Note, this is friction around the dowel NOT along it.  The line from the constrictor was then wound around the dowel and exactly as predicted, the gripping force rapidly amplified within just a few turns.  Even tying this construction on a glass bottle it was easily possible to add turns until the small holding force from the Constrictor had been amplified to a level sufficient to grip the glass.

Of course, this is grip tangential to the dowel and this hitch must create grip in line with, i.e. along the dowel.  This was easily achieved by driving a pin into the dowel and turning the cord 90 degrees around the pin so that the gripping force now ran along the dowel.  Of course, we cannot rely on banging pins into object just to make our knots work - this knot would have to provide its own means of turning the direction of grip through 90 degrees.  The solution is of course an easy one.  The second line coming out of the constrictor was simply wound in a counter spiral through the first and the two ends joined.  In this manner the two coils work together to turn the gripping force through the necessary 90 degrees.

Finally the force or tensioning engine.  As the loading force is applied to the knot, the first turns start to open up - concertina fashion.  In doing so, the cord is forced to lengthen and the force of creating this stretching is transmitted through the cord back into the knot itself.  This linear tensioning of the knot cord massively increases the coefficient of friction and is transmitted right back to the anchor Constrictor further tightening it and increasing the anchor grip being made available to the gripping engine coils.  The more elastic the cord is or in cases of very low friction, several of the turns might be forced to open in order to generate sufficient line tension within the knot.

And there you have the prototype "KC Hitch".  It is highly functional, it can grip glass even with Spectra cord and it can hold firm onto a taper.  There is a limit to the angle of taper this knot can hold and this limit happens when the diameter of the taper narrows faster than the line extension caused by opening the loops.  When this angle is reached, the knot will never hold and will always simply run off the taper.  However, at angles below this critical angle, line tension is increased as the loops open up and given sufficient turns, any surface can in theory be held.

Optimisation

The prototype "KC Hitch" is highly functional, but as a working knot it is a real pig to tie and so is unlikely to ever move into mainstream use.  It was notice during trials with various cords and binding surfaces, that movement of the last two coils rarely occurred and that the tensioning effect caused these coils to lock tightly onto the load surface.  This meant that the coils could act as their own anchor thereby dispensing with the need to tie a Constrictor around the load.  This proved to be the case and led not only to a simplification in the construction of the knot but also to an extremely simple method of tying it.

Here then is the final field ready "KC Hitch"

Take the line and start with a working length sufficiently long to make the required number of turns around the load.  For high friction loads four is sufficient for low friction (glass, polished wood etc) eight may be necessary. 
Hold the cord against the load with your thumb

 and wrap the working end around the load in an open spiral ( say three or four turns while you experiment with this hitch).
 
Then start wrapping the cord back over the first spiral so the cords cross front and back of the load.
 
When you get back to the start, tie the ends together using whatever knot you prefer, even the Reef will do here as its only function is to hold the ends together.

 slide the coils up together and take out the slack.

 Apply tension to the load and the first coils will open and tighten up the knot.  If there is a lot of slack taken up, then close up the loops again and remove the slack before reapplying the tension.

The objective is to have at least the last two turns (the anchor turns) remain closed under full load conditions.


Observations

This knot uses leverage to apply tension and that leverage is very powerful.  I have seen samples where 500# Dynema has bitten into the surface of a glossy dowel - it did not slip, but it sure bit in. As a consequence, you need to be sure that you are using rope or cordage of sufficient strength to take these loads.  This needs to be considered as a potential weakness of the knot as the leverage is capable of amplifying the applied load to levels in excess of the breaking strain of the cordage used.  You might find it advantageous to tie the knot to the load in one cord and then use a different rope to haul with.

The second observation may serve to stimulate future studies, and that is that there is a definite relationship between cord diameter a load diameter and how many coils open during the loading phase.

Conclusion

So there you have it - is it a new knot?  Does it work in all applications you test it in?  What is your considered opinion of its structure and working parts?  Do you feel that the analysis and process was flawed in any way?  When does the knot not work and what safety concerns do you have?

Is it now possible for members to give this knot a critical Peer Review as this is surely the prime value of the combined expertise of the people gathered around the IGKT.

[bridog]

I'd say ABOK1755.  In #1756, however, it is suggested that not just crossing the line over itself, but actually reversing directions, is a possibility.  I would expect line reversals of this nature to create a larger mechanical advantage as the cord is working against itself.  In #1755 and your images, the line will work to rotate the entire body in a given direction, and the only point that counters that rotation is the final holding knot.  If instead of just laying the cord over itself, we actually pass it around itself and reverse it multiple times around the pole, it will have nowhere to go whatsoever.  Indeed, the final knot still supports the totality of the force, but each wrap would improve gripping ability much more without causing rotation. 

As to actual physical measurements and comparisons, I have only made several (and nothing worth publishing), but I will say I'd still expect directional reversal and line crossing to amplify grip much faster than unidirectional coils.  The nature of force amplification per coil is also questionable; if it were truly exponential, one would expect you'd be able to use this to break rope after several coils --- alas that this were true.

[squarerigger]

Hi Derek,

A splendid try - it does look somewhat like 1755 in ABOK but I think your construction method is a little different in that only one side of your line is tightening over the other all the time, tending as bridog said, to twist the object to which this is gripping.  We use a similar construction to 1755 (ABOK 1758) when applying a selvagee to a shroud for tuning the rigging and we take that loop (you could still use a tied piece of line) and wrap it over itself in alternating turns - first the left is passed over the right then the right is passed over the left, so that the load is shared equally (by load here I mean the tension developed along each leg of the selvagee) by both legs and the friction applied is also shared without having to shove each succeeding turn up against its neighbor as you have done.  However, I do admire your method - it looks very thorough (of course we still have to hear from Dan L on all the points/line/objects it does not work for) and easy to replicate.  I would like to hear of more of this type of application!

Lindsey
PS - I wonder if bridog's method of reversing (I take this to mean the formation of elbows in the pair of lines) the lines will result in point loads instead of spread loads, where the spread loads of your wrap or the wrap I have suggested, maximize friction?  Bridog - I agree that it is difficult to see exponential growth here!

[DerekSmith]

Hi guys,

Thanks for the responses.  If it is 1755, then I suppose it just goes to show that nature has a way of inventing the same thing repeatedly?

There are a couple of the issues you raise that I would like to comment on.

First reversal;  if by this it is meant that effectively each cord stays on 'its own side' and when the cords meet, they cross and do a 180 degree turn.  Then I would say that it is critical NOT to apply reversal.  When a cord changes direction by being turned around another, considerable tension is lost depending on the radius and extent of the turn.  In the proposed hitch there is not a single reversal within the knot (i.e. after the tie off) and this allows tension generated in the force engine component, to flow through the cords right to the anchor coils.  You can test this by seeing how hard the cord in the end anchor coils goes when tension is applied.  This compresses the cord in the anchor section onto the load object without placing the coils under any 'in line' tension at all.

If however, by reversal you mean 'priority' i.e. that at the first crossing A is over B, then in the next crossing B passes over A, etc. etc.  Then this was the form of winding used in the 'prototype' knot.  However, it is cumbersome to tie when you only have access to one end of the cord or the load is long, and in trials, the single priority wrap method showed no loss of performance.  In fact, leaving one cord always on top may aid the transfer of tension back to the anchor coils, however, there was no marked indication of this.

Second - the concept of 'Force Amplification per Coil' is wrong.  Grip is amplified by the coils, not force.  On a slick surface such as the one shown in the example, especially with a slippery braid, getting any level of grip is hard.  However, once you have achieved some level of grip, tension in consecutive coils builds this grip at the rate of Euler's constant to the power of the coefficient of friction times the total angle of contact in radians.  Tension in the cord however, may be massively increased by the mechanical advantage caused by the coils opening in the 'force engine' part of the knot.  Although this is the purpose of this stage, it is also the prime weakness of this knot that it can either break the cord or damage the surface of the object being held if low friction and high dragging force causes too many coils to open.  Note: it is not the number of coils that influence this, it is the number of coils that are opened that causes the mechanical advantage.

Finally, the importance of 'shoving the coils up against one another'.  The 'diamond' form of the knot works well, and can be used if you do not want the knot to be allowed to develop too much tension through mechanical advantage.  However, the mechanical advantage is at its highest when the angle between coils is zero degrees, so closing the coils (and removing the free cord generated) gives the oportunity for greater tension to be created within the knot.  If too much tension could be an issue in use, then deliberately leaving the first two coils in 'diamond formation' will dramatically reduce the effectiveness of the the 'force engine' component.

Derek

[drjbrennan]

I did some experimenting with the 'KC Hitch' today, on broomsticks and aluminium bars, and from the first try it performed very well. Easy to tie and cast off, with little adjustment needed for a ratchet-like grip. It's definately in my list of favourites now. 

[KC]

Sir Derek i'm honored!
And this is my 4th try to respond to this due to computer type problems (and length of response?).  Also have much less time available here; and pretty much everywhere!

i think this strategy/ mechanical configuration is less like knots we know better(and thereby their mechanics); and more like a Chineese Finger Puzzle or loose braid/ weave; though following same principals.

Because we pull at both ends of the laced line; instead of 1 like in a Constrictor; we immediately have half as much power/ line tension loaded into each leg and curve.  In a Constrictor, full force is initiated on each/ the single leg of it's turns.  i've looked at rope as just a  hose device that carries/ conducts tension force flow down it so much; we even call an electrical extension cord an 'electric hose' around here!

In this imagery of a rope that is just a hose loaded with not the push of water pressure force; but the pull of tension force; i see a Constrictor's pull on it's host/ mount/ bundle / load is best at perpendicular to this host; for that places the force of it's bends inline/ not perpendicular to the initiating Standing Part's pull.  This "KC" hitch is best pulling inline with the host/ mount/ load for this also; pulls most inline with it's bends.  It is just that the bend in Constrictor is near Zer0(comes back to Standing Part / Initiating Force, and in KC it is near 180 and extends from there further away from Standing Part/ Initiating Force.   So Constrictor pulls perpendicular to mount, and KC inline; reverse scenarios of force flow, giving reverse useage mechanics IMLHO.

Both even give fairly balanced pulls to this direction too.  A Constrictor comes in center mast to it's turns, and then flows the rope/ line of force to one side, then the other; before force terminates in center (of hitch) once again.  KC/ Finger Puzzle; pulls from both sides of host / mount evenly; then cascades residual forces back and forth fairly balanced; but all ways a net of inline with Standing Part as Initiating Force.

Also thus; Constrictor gives an intense compression/ push inward in one area perpendicular to host/ in a strianght line across it's narrow axis.  While KC gives a sock/ glove of pull along the length/ long axis by gripping mass of the host; once again opposite effects.  A Constritor's tensile strength is probably, mostly found to be dependant/ degraded by size of host arc i'd think; while KC's strength is mostly the first / most loaded arc of 1 leg of line bending the Standing one; to make both of them a Standing Part/ Initiating pull into the next bend/ arc.  Making both ehnds of KC strong eyes to hook directly to Standing Part; without bend in most loaded part of line as most ineeficient point; and more the long arc taken around host; shared betwen 2 lines of pull.  It would thus be more of a softly bent basket pull, rather than sharper/ bend around self single leg pull.





In certain ranges to increase leverage on bend/ arc we must use 1 strategy to get maximum tightening and in it's opposite arc range; likewise the opposite tightening strategy; i think.  In a bend/ arc close to Zer0 degress/ U shaped; we have 3 positions.  2 ends of the bend (as a mechanical unit) and the bend itself.  In this instance/ range; to tighetn more; we anchor one end and pull from the other as input; to get an output of higher leverage at the bend.  As in a 2/1; only isde the microcosm of a knot/ lacing.

Notice if we input pull at the bend; instead of the end in a Zer0 degree range of bend; we are just dividng our pull between the 2 ends; at inline angle to the pull.  Inline angle to bend being most direct direction/ least leveraged; so am just dividing unleveraged pull.  So this is backwards / reducing my pull effort; not increasing it; like pulling from 1 end give output (at bend) increase above effort input. 

If however we have a bend/ arc that is of opposite Nature; 180 degrees range/ folding away from Standing Part/ Initiating Force (instead of folding towards SP); then we must apply opposite strategy for most tension i believe.  Of our 3 positions, of 2 ends and the bend; we still anchor 1 end; but now anchor the other/ or place on load/ host etc.; and input force at the bend; to recieve it at the end (or really both ends) as in Brion Toss's sweating more purchase out of a line.  If we pull at the end of this configuration (and not the bend) like KC Hitch does; then i believe we would be losing potential leverage force;as we are inputting at end; when we should input force at bend; and recieve output force at end(s).  All theoro-correctable of curse; but the way i see it!

Also, this run to first bend of KC being a contacting one; engages friction; so their is more friction reduction of input/ Standing Part pull going into first bend, than the force in Constrictor at first bend; even if we compensate for the 1/2 force of 2 legs of support in KC Hitch i'd think.


KC Hitch shows 3 bends/ Round Turn as very powerful strategy once again; even as the spiral is elongated out to pull linearly/ inline with host mount and not perpendicularily.  If just 1 bend, would push host perpendiculairly.  2 bends but not inline; would twist host/ mount.  But 3 bends gives 2 pushing on host in 1 direction, and a center bend; pushing balanced between those 2 points for a more 'stable' grip/ trap on line.  Especially noticeable if we take our friction hitch from a non-flexible host mount (pipe, spar etc.) and place it on a flexible host/ mount (rope grip etc.); that can bend in the grip of the KC or similair hitch.  Subsequent turns becoming more superfluous; unless mount very slippery i think.

Ummmmmmm i guess my turn is over!

[DerekSmith]

I'd say ABOK1755. 

In "The Ashley Book Of Knots" p299 ref #1755, Ashley depicts the following construction:-



and says of it:-

"1755.  A cross-lashed strap made fast in the rigging, to hook a block to.  Shakespear terms this method of lashing (differently applied) - cross-gartering"

From the diagram, it is clear that the #1755 described by Ashley is essentially the structure depicted in step four of the production of the KC Hitch.  Clearly mariners of the time of Shakespear knew of the value of a cross-gartered lacing and its ability to hold laid rope onto vertical (timber) rigging.  However, its structure and the description lead me to believe that its inventors and users had little idea how tantalisingly close they were to a knot of far greater potential (I can but assume that in  the materials they were working,  the #1755 structure was none the less more than addequate) and for the reasons I will explain, I feel that the KC Hitch remains a new knot, worthy of its own name.

Of the number of hitches recorded by Ashley involving lines which are multiply 'dogged' (i.e. wrapped around the spar), they are all depicted in use in the open (i.e. gartered) configuration (1751, 1752, 1753, 1755, 1756, 1758).  Because of this all of these knots have a force vector directly in line with the load, applied through the knot to the top of the knot (the anchor part of the KC Hitch).  Doubtless, in the materials of the day, the coefficients of friction were so high that there was no need for a anchor section, nor need to amplify this grip.  The gartered body of the knot was more than sufficient to produce all the contact grip necessary for the job.  Try this today with braided nylon on chrome tube and the #1755 will slide straight off!!

However, those mariners were tantalisingly close and had they needed the additional grip, doubtless one of them would have stumbled across the process of closing up the lacings before applying the load, and of noting that sufficient coils must be applied to ensure that the last two coils must always remain closed to keep the knot in place.  But they didn't need the grip and the "Cross Garter lashed strap" is as far as they got.

So, although the 1755 is close - it is half way there - it cannot perform the same feats of griping made possible by the enhancements which make the KC hitch the knot that it is.

Derek

[DerekSmith]

ABOK #1755 was the closest contender for the KC hitch.

Now I have come across a bend based on the same structure -- the Albright knot.

Interesting that it works very well when the KC cord is much lighter than the other.

http://www.thaifishingguide.com/fishtechequip/techniques/knots/albright_knot_steps.html

[Dan_Lehman]

Now I have come across a bend based on the same structure -- the Albright knot.

One of the continual problems with the presentation of angling knots is that the
actual, final geometry is obscured (and I guess that this is partly attributableto the
author not knowing what it is!).  The part of the Albright knot that superficially
resembles the KC knot is not at all present upon completion--the wraps of the
SPart--which is loaded w/o sharing with the end, in contrast to the KC knot--
will pretty much straighten out and cast all of its wraps into the outer, overwraps,
as your cited presentation shows (though it obscures the SPart, lying beneath).
The Albright is akin to the Blood knot, to Common Whipping.

--dl*
====

[Bob Thrun]

I called this a French Prusik in a 1967 newsletter article that got expanded into a book in 1973.  Bill Plummer's highly mathematical analysis (1963) of it is included in the 1973 book.  Paul Hasluck called it a cross lashing in his 1905 book. The US Army rigging manual calls it a telegraph hitch, used for gripping posts while lowering them into holes.  The name "telegraph hitch" should give some hint of its age. The name dog and tails is used for a variant where the closing knot is at the unloaded end of the hitch

[DerekSmith]

The following information from icicleuk.com http://www.icicle-mountaineering.ltd.uk/skills.shtml shows the only image I have found for the French Prussic which is often referred to as a French Autoblock / Kleimheist Prussic (presumably stemming from the similarities of these two knots).  Neither of these two knots have the garter hatch crossings of the KC hitch.

  Kleimheist Prussic
The kleimheist is often the most effective prussic for locking on a rope, and is frequently used for a safety back up on an abseil. The only disadvantages to this knot are that it can lock too well, and be hard to undo, and also all the load weight is on one piece of 7mm prussic cord in a fall situation.
  French Autoblock Prussic
This prussic looks similar to the kleimheist, apart from both end loops are clipped into the karabiner, so in event of a fall, the weight is distributed better on the prussic. This knot is very useful in a crevasse rescue hoist pulley system as it can be easily loosened, and can lock off automatically.

However, the Tress Prussic on the same site does indeed show the KC structure.

Tress Prussic
The tress is the best prussic to use on wet or iced up ropes, a it squeezes the rope to create friction when under load, rather than constricting on itself like a normal prussic. This enables it to open more when not loaded, so it can easily be slid over lumps of ice that have formed on the rope.

However, as Prussic knots are usually put on using a loop, unless this one is put on over the end of the rope, I cannot see how it would have been created using a preformed loop.

Paul Hasluck's description of cross hatchings is reminiscent of ABOK #1755 and either #1755 or the KC hitch would make an excellent 'telegraph hitch'.

As for the 'dog and tails'.  That is an excellent enhancement.  The weakest part of the KC and the KC Sling Hitch are the closure point.  The 'dog and tails' concept removes this weakness to the least stressed part of the knot, i.e. as you put it Bob, to the 'unloaded end' (it's not really unloaded but it is the least loaded).

So, the ultimate way of tying this knot for strength might be as follows:-

Middle the cord and form a cats paw.  Place the cats paw over the hauling device (hook, bina, chain ??)
Take the two ends and proceed wrapping the gartering around the spar, load etc.
When enough wraps have been applied to use available friction, close up the garter diamonds and knot the ends tightly (consider a slipped knot for ease of later release).
Apply the load and check that at least the last two coils near the closure knot do not need to open to achieve the necessary grip.

I think this one may well be the hitch to beat !!

Of interest, I have never thought of the KC hitch as being of any use in a cord on cord situation because it deforms the main cable in its attempt to gain purchase.  However, having seen it under test, it is acting very much as the plaited double performs, shedding load by latteral compression rather than by constriction, so again the use of this structure in the 'Prussic knot' mode could well prove to be very strong.

http://cmru.peak.org/Rigging/dnt.htm


Derek

[knudeNoggin]

'prussic' => 'Prusik' ,

'kleimheist' => 'klemheist'

And I think this has been referred to before, but a good article about fricion hitches (in HTML of PDF) is:
www.google.com/search?q=cache:39oMYLWibwcJ:www.treebuzz.com/pdf/climbing_hitches.pdf+valdotain+tresse&hl=en&ct=clnk&cd=1&gl=us

*knudeNoggin*

[Bob Thrun]

The name "French Prusik" is ambiguous.  I have found at least six hitches by that name.  My 1967 article and 1973 book may have been the first publications to describe the "French Prusik".  I found the name used in a 1961 rescue report with no description.  Heinz Prohaska asked me where I got the hitch and if the French actually use it.  I do not know.  The name was in common use by thos from who I learned, but they did not know its origin.  I bought about a dozen pre-1960 climbing books.  None of them have it.

Michael Noonan, in his 1997 book "Climbing Knots for Righties and Lefties", used the name for what I would call a Helical Knot.  So did Richard Hopkins in his 2003 book "Knots".

The most recent French Prusik is from Alpine Caving Techniques by Marbach and Tourte.  It amounts to a single-strand Bachmann. The name was a choice by the translator.  The original French edition calls it a "noeud autobloquant" with carabiner.

The name "Autoblock" for another hitch is a bad translation. The generic French term for these hitches - Prusik, Bachmann, Hedden, etc. - is "noeud autobloquant".  I can imagine an English-speaking climber asking a Frenchman "What do you call that?"  The French call it a "noeud Machard".  I would like to research the origin of "autoblock", but I have not been able to find someone with a 30-year stack of climbing magazines and books.

[Tom]

Head still not working very well, so please forgive me if this has all been covered in some of the stuff that I didn't follow above, but the knot is familiar to me too in a climbing context, though I have never had a name for it beyond 'plaity Prussik' which was my own coinage, but may have spread beyond the confines of my own small band of belayers (and I certainly didn't make up this knot myself!). My available library extends only to Geoffrey Budworth's 'Need to Know? Knots' that I bought last week - in a bid to stave off insanity - from the excellent Postscript Books (http://www.psbooks.co.uk/BookDetails.asp?Code=38498&pg=Title+Search+Results&ur=SearchResults%2Easp%3FTitle%3Dknots?pgn=1%23Nav38498) for a reasonable ?4.99, but  - rather gratifyingly - within he illustrates the same fellow (though secured with a shackle and tied in a sling) on p.162-3 and he calls it the 'slide-and-grip hitch (end-loaded, one-way)'. A catchier title than KC hitch? It will certainly grasp the smooth aluminium of my crutches even when tied in very slick 8mm cord. Budworth, too, suggests the Chinese finger trap comparison.
All these 'Prusik' hitches get their names swapped around fairly frequently, dependent on with whom one is climbing, and perhaps the Guild is the place to firm up the nomenclature once and for all. They each bring something different to the mix, and perhaps a chart of their most appropriate applications might be formed (once one has settled on the names...)
By the way, can anyone tell me how Dr Prussik actually used his knot for repairing instrument strings, and does it work? Does anyone use it thus today? He has spawned a verb in mountaineering, but I've not heard of him in a musical context at all.
I have been very much enjoying another doctor's work - Asher's Bottle Sling! A variant on Prussik Classik (will this term catch on? You heard it here first! Illustrated by Budworth (ibid. p.94-5) which I had not met before, but seems to top the fiendish Jug Sling that I can never remember how to tie.
And to conclude my ramble: Derek, does your interest into the numerous friction hitches extend to the range of slide and grip hitches of the Midshipman's variety? I use these a lot for all sorts of tightening/guy-rope type functions, and it seems to me that almost anything asymmetrical (about the point where the working end meets the standing part) will do, so long as it doesn't shake free - exploiting the opposite effect to your lovely symmetrical hitch above. All the Tarbuck/Midshipman/tautline/rolling hitch family rely on this kinking of the rope-that-is-to-be-gripped, and my own favourite (nameless?) brings the working end back over the SPart in the final half-hitch to run parallel to itself, as it were, for elegance - I'll take a picture if there's the interest, but it requires stair-use, so that's for another day...

[DerekSmith]

I originally 'designed' the KC Hitch as an enhancement to the the standard hitches in simplicity of tying and maximised grip on tapered or slippery objects.  The gartered lacing had been utilised many times before but always open, presumably to spread load, never closed as in the KC where the design was that as line tension increased, the lacing opens up with massive leverage advantage and creates huge grip pressures in the remainder of the hitch.

I had not thought to use it on other cords, indeed when I tested it in the manner of a 'prussic' I regarded it as very poor.  In fact, apart from tying a very light line onto a very stout mono, I still do not favor the lacing or braid prussics over the significantly better performing 'wrap' style prussic (I use 'prussic' here in the sense that one would use nescafe, biro or hoover to denote a generic class of objects -- in this case, a knot formed out of lighter line to create an adjustable grip on a heavier line).

Re Bob's post; has anyone else notice that there seems to be a trend in the evolution of the use (and therefore popular meaning) of the term 'prussic' - vis; 'a hitch' is used to tie cord to a solid object, while 'a prussic' is used to tie cord to a thicker line (sliding adjustment incorporated).

Tom; re other friction hitches.  It was study of how other hitches worked that led me to consider that they were nearly all flawed? because they mostly took the load straight to the back of the knot which tried to force the knot to slide.  When a knot did slip, it was because this loading was pushing the knot down the thing it was fixed to.  For a knot to escape this flaw, it had to have a 'safe anchor' removed from the incoming load.  My experimentation started with a constrictor as this 'safe anchor', but I soon realised that simple wraps were more than adequate once the exponential friction law of multiple wraps got put into gear.