Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

How is this mystery dovetail done?

11 views
Skip to first unread message

Chita

unread,
Oct 12, 2005, 9:26:53 PM10/12/05
to
The picture looks familiar but I can't recall the details. I think
there was a "wood puzzles" book which included this novelty joint. (Was
it a book sold by Woodcraft?) Anyway, it's a curiosity. Check it out
and tell all.

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=6976867662

Lowell Holmes

unread,
Oct 12, 2005, 10:21:37 PM10/12/05
to
"Chita" <chita...@gmail.com> wrote in message
news:1129166813.5...@f14g2000cwb.googlegroups.com...
Roy Underhill shows how to make the dovetail in his book "The Woodwright's
Apprentice".
I couldn't begin to tell you how to do it without illustrations. :-)
I've never attempted the dovetail.

--
Lowell Holmes


Morris Dovey

unread,
Oct 13, 2005, 12:57:14 AM10/13/05
to
Chita (in 1129166813.5...@f14g2000cwb.googlegroups.com) said:

I'd have to guess that there isn't much wood between the two dovetails
:-)

--
Morris Dovey
DeSoto Solar
DeSoto, Iowa USA
http://www.iedu.com/DeSoto/solar.html


mywebaccts (at) PLUGcomcast.net

unread,
Oct 13, 2005, 1:14:02 AM10/13/05
to
I've seen puzzles like this before that were accomplished by soaking the
wood until it got soft and pliable, positioning it, and then letting it
dry. The one I remember is a wooden ball inside a cube that has square
holes through each side (but just slightly too small for the ball). The
cube is soaked and then "squeezed" around the ball.

Don't know if this is the trick for the dovetail, but that would be my
guess.

Jack

mywebaccts (at) PLUGcomcast.net

unread,
Oct 13, 2005, 1:30:47 AM10/13/05
to
Ah .. wait! I was staring at the picture and I just figured out a way
that you could do this with solid pieces of wood without needing
'pliable' wood.

In the picture, there are brown and red woods dovetailed together. If
you look at the red piece, imagine that the small strip of red wood at
the top of the puzzle is in fact just that ... a small strip of wood.

That is, the two sides of the red piece are pretty much what you see,
but the section between the two dovetailed brown pieces isn't there ...
just a triangular "rod" connecting the two sides of the red piece.

Then ... you can easily "pivot" the red piece up around the brown dovetails.

The red rod connecting the two sides would need to be cut inwards to
allow for the pivot, as would the bottom ends of the red piece that butt
up against the brown piece.

Can you envision what I'm describing?

Jack

Mike Marlow

unread,
Oct 13, 2005, 7:36:46 AM10/13/05
to

"Lowell Holmes" <lh...@hal-pc.org> wrote in message
news:Rwj3f.23810$wm3.18096@trnddc01...

I've seen it as well. Both pieces are cut on a diagonal but when put
together they look like a dovetail - though they really aren't (if memory
serves...). I'd have to find the web page that breaks it all down too. I
couldn't explain it any better right now than what I've just said.

--

-Mike-
mmarlo...@alltel.net


Toller

unread,
Oct 13, 2005, 9:36:03 AM10/13/05
to

"mywebaccts (at) PLUGcomcast.net" <"mywebaccts (at) PLUGcomcast.net"> wrote
in message news:v46dnbFxC5C...@comcast.com...

> Ah .. wait! I was staring at the picture and I just figured out a way
> that you could do this with solid pieces of wood without needing 'pliable'
> wood.
>
> In the picture, there are brown and red woods dovetailed together. If you
> look at the red piece, imagine that the small strip of red wood at the top
> of the puzzle is in fact just that ... a small strip of wood.
>
> That is, the two sides of the red piece are pretty much what you see, but
> the section between the two dovetailed brown pieces isn't there ... just a
> triangular "rod" connecting the two sides of the red piece.
>
> Then ... you can easily "pivot" the red piece up around the brown
> dovetails.
>
> The red rod connecting the two sides would need to be cut inwards to allow
> for the pivot, as would the bottom ends of the red piece that butt up
> against the brown piece.
>
> Can you envision what I'm describing?
>
That was my thought as well, but I don't think the other end would clear
unless it was cut really really loose; and it doesn't look that way.


brianlanning

unread,
Oct 13, 2005, 9:58:05 AM10/13/05
to
Yeah, there was an episode that talked about this joint. If you look a
the end grain of the piece with the tail cut in it, it's a square cut
which creates the illusion that it's a square cut all the way back to
shoulders like a normal dovetail would be. That's impossible though.
What's really going on is that the sides of both dovetails are beveled
so that where the dovetail attached to the rest of the board, the sides
of the tail are at a 45 degree angle. The board simply pushes straight
in from the back. Between the two dovetails is sort of a V-shaped
valley with a matching one on the other piece. It's difficult to
describe, but easy to understand once you see it.

brian

Gordon Airporte

unread,
Oct 13, 2005, 1:19:22 PM10/13/05
to
I saw a Woodwrght's shop episode where Roy made one of these. Unlike the
other trick dovetail (where the tails appear on all four faces of the
joined pieces) this one required heavy clamping to bend the pieces when
it was assembled. Other than that I can't recall any details. It is
really secure, though - he used it for a mallet head.

Bruce Barnett

unread,
Oct 14, 2005, 7:31:53 AM10/14/05
to
"Mike Marlow" <mmarlo...@alltel.net> writes:

>> > http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=6976867662


>
> I've seen it as well. Both pieces are cut on a diagonal but when put
> together they look like a dovetail - though they really aren't (if memory
> serves...).

That's a different puzzle. I doubt this one was by cutting on diagonals.

--
Sending unsolicited commercial e-mail to this account incurs a fee of
$500 per message, and acknowledges the legality of this contract.

Mike Marlow

unread,
Oct 14, 2005, 3:20:44 PM10/14/05
to

"Bruce Barnett" <spamhater113+...@grymoire.com> wrote in message
news:dio4v9$rlp$0$208.20...@netheaven.com...

> "Mike Marlow" <mmarlo...@alltel.net> writes:
>
> >> > http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=6976867662
> >
> > I've seen it as well. Both pieces are cut on a diagonal but when put
> > together they look like a dovetail - though they really aren't (if
memory
> > serves...).
>
> That's a different puzzle. I doubt this one was by cutting on diagonals.
>

Could well be. Was pretty cool the way the effect worked though.

--

-Mike-
mmarlo...@alltel.net


Bruce Barnett

unread,
Oct 14, 2005, 5:08:17 PM10/14/05
to
"Mike Marlow" <mmarlo...@alltel.net> writes:

>> That's a different puzzle. I doubt this one was by cutting on diagonals.
>>
>
> Could well be. Was pretty cool the way the effect worked though.

I haven't seen the actual mechanism. But one poster said clamping was
needed, so perhaps it's a combination of sliding and clamping.
So you may be right....

fredf...@spamcop.net

unread,
Oct 14, 2005, 7:04:27 PM10/14/05
to

mywebaccts (at) PLUGcomcast.net wrote:
> I've seen puzzles like this before that were accomplished by soaking the
> wood until it got soft and pliable, positioning it, and then letting it
> dry. The one I remember is a wooden ball inside a cube that has square
> holes through each side (but just slightly too small for the ball). The
> cube is soaked and then "squeezed" around the ball.
>

I thought the ball and cage was typically done by carving both
from the same block of wood, carving the ball _inside_ the cage.

--

FF

Greg O

unread,
Oct 14, 2005, 7:50:13 PM10/14/05
to

--<fredf...@spamcop.net> wrote in message
news:1129331067....@z14g2000cwz.googlegroups.com...

Not wood related, But I once worked for a CNC machine shop. To prove what
the machines could do one of the owners machined the "ball in the cube" out
of aluminum, about 2 inches square. It was surprising to see the end result.
The ball was smooth, and measured with in .001 of round. All machined out of
one piece, the ball inside the cage.
Greg


Greg O

unread,
Oct 14, 2005, 7:55:15 PM10/14/05
to

"Mike Marlow" <mmarlo...@alltel.net> wrote in message
news:9531a$434e4680$a2270060$21...@ALLTEL.NET...


I have seen it too, just as you describe.
The end result is an illusion, so to speak. Our minds are trained to think a
typical dovetail, but the wood is cut in a way to simulate a dovetail, when
it is really a form of slip joint.
Greg


bri...@all.costs

unread,
Oct 14, 2005, 8:06:48 PM10/14/05
to

don't confuse 'em, fred...

bri...@all.costs

unread,
Oct 14, 2005, 8:09:06 PM10/14/05
to
On Fri, 14 Oct 2005 18:55:15 -0500, "Greg O" <goo...@cableone.net>
wrote:


it's a good illusion, then. I can't see how it could be assembled, no
matter what it looks like inside.

what are the dots for?
why won't the ebay vendor let us see the underside?

Mike Marlow

unread,
Oct 14, 2005, 8:41:51 PM10/14/05
to

<bri...@all.costs> wrote in message
news:h2i0l1t7s61ckjbde...@4ax.com...

OK - I'm going to give this a shot from memory. What you see is a box with
apparently two dovetails - on adjacent sides. What's really there is a
sliding joint. Imagine that the dovetail end you are seeing on one side
runs to the other one you see. What makes the illusion work is that the
ends of the dovetails are cut on an angle - even with the plane of the side
of the box. If you looked at the box straight on, but looking into the
corner instead of at one of the sides, you'd see the dovetail profile with
the ends cut off-square. The top has the dovetail slot to match. It slides
together.

Damn - I don't know if that made one bit of sense.


--

-Mike-
mmarlo...@alltel.net


Mike Marlow

unread,
Oct 14, 2005, 8:48:22 PM10/14/05
to
I posted a really crummy Word doc drawing of how this works on the binary
group.

--

-Mike-
mmarlo...@alltel.net


A. D. Coby

unread,
Oct 14, 2005, 9:13:30 PM10/14/05
to
One possible way to make this joint would be for the hidden
cuts to be circular arcs. The sort of thing that would result from
a rotated (turned) piece which then has been cut with flat sides

The axis of rotation would be along a line on the surface between
the labels A and B.

If so then there would be arc shaped cut inside (under) the
arms of the T. This view is not given in the pictures.


mywebaccts (at) PLUGcomcast.net

unread,
Oct 14, 2005, 9:15:45 PM10/14/05
to
Well, you could do it that way. But I saw it with a nice hardwood ball
(oak or something) inside a cube made from some different wood.

Bruce Barnett

unread,
Oct 14, 2005, 10:32:51 PM10/14/05
to
fredf...@spamcop.net writes:


I've seen one in Popular Science from the 1960's that had an arrow
going through a glass. The drinking glass has two holes in it - the
same diameter as the shaft of the arrow. The Arrow head and feathers
(fletching) were twice the width of the glass hole.


This one was done, AIR, by soaking, squeezing the head in a clamp,
pushed through, an then soaked again do it swells. I think it was
balsa wood.

alexy

unread,
Oct 15, 2005, 12:08:44 AM10/15/05
to
"Mike Marlow" <mmarlo...@alltel.net> wrote:

It does. But it falls into the category of "if you don't know the
answer to the question that was asked, give an answer you do know to a
question that was not asked" <g>. I like the puzzle you are referring
to, and its deceptively simple solution. But I'm pretty sure that is
not what the OP is asking about. Look again at the cited ebay page,
and see if there is any way you can see the puzzle you are talking
about morphed into this one.

--
Alex -- Replace "nospam" with "mail" to reply by email. Checked infrequently.

A. D. Coby

unread,
Oct 15, 2005, 12:20:36 AM10/15/05
to

Another possible way to make this joint is to make the strip of wood across the
top of the joint between A and B thin enough to be easily bent. Then the piece
which is labelled A B would be flexed open and the moved pass the tails in the
other piece and then first piece would then be bent back closing the joint
around the second piece.


Mike Marlow

unread,
Oct 15, 2005, 7:25:55 AM10/15/05
to

"alexy" <nos...@asbry.net> wrote in message
news:3001l11nunmhrku1d...@4ax.com...

>
> It does. But it falls into the category of "if you don't know the
> answer to the question that was asked, give an answer you do know to a
> question that was not asked" <g>. I like the puzzle you are referring
> to, and its deceptively simple solution. But I'm pretty sure that is
> not what the OP is asking about. Look again at the cited ebay page,
> and see if there is any way you can see the puzzle you are talking
> about morphed into this one.
>

Argh! You're absolutely right. My bad. A mind is a terrible...

--

-Mike-
mmarlo...@alltel.net


John Briggs

unread,
Oct 15, 2005, 10:57:20 AM10/15/05
to


Here is the definitive source:

It's from E.M Wyatt's "Wonders in Wood" (1946).

Jerry Slocum's and Jack Botermans' "Puzzles Old & New" (1986) provide this
information and an illustration.

John

charlie b

unread,
Oct 15, 2005, 1:44:19 PM10/15/05
to
Chita wrote:
>
> The picture looks familiar but I can't recall the details. I think
> there was a "wood puzzles" book which included this novelty joint. (Was
> it a book sold by Woodcraft?) Anyway, it's a curiosity. Check it out
> and tell all.
>
> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=6976867662

Think "slight taper", "notch" and "rotation".

What I think is a solution, posted in alt.binaries.pictures.woodworking
as '"Impossible Joint?" - solution?, should make my hints a lot
clearer.

charlie b
(If this seems like the solution to you
send half of the $2.95 US I saved you
to your favorite charity)'

bri...@all.costs

unread,
Oct 15, 2005, 4:15:39 PM10/15/05
to
On Sat, 15 Oct 2005 01:13:30 GMT, "A. D. Coby" <dan....@artifex.com>
wrote:


this is so far the only explanation that fits the situation...

fredf...@spamcop.net

unread,
Oct 15, 2005, 5:09:13 PM10/15/05
to

Well, that is much cooler than the other one.

--

FF

fredf...@spamcop.net

unread,
Oct 15, 2005, 5:15:06 PM10/15/05
to

Now who is confusing whom?

--

FF

Limey Lurker

unread,
Oct 15, 2005, 5:21:09 PM10/15/05
to


Yep, that's how it's done!

charlie b

unread,
Oct 15, 2005, 7:51:19 PM10/15/05
to

But where's the fun in that? I mean it doesn't even use
- a lathe
- a block plane
- a shoulder plane
- block rabbet plane
- skew chisels
- bench chisels
- paring chisels
- micro belt sander
- drill press
- 20 special router bits
- a precision router lift
- precision router table fence
- oscillating spindle sander
:
:
:
Where's the fun?

Guess there really is more than one way
to skin a shop cat push stick. Who'd've
thunk it?

bri...@all.costs

unread,
Oct 15, 2005, 7:56:46 PM10/15/05
to
On 15 Oct 2005 14:21:09 -0700, "Limey Lurker"
<steve....@tesco.net> wrote:


that particular joint would have to be flexed a _lot_....

Doug Warner

unread,
Oct 15, 2005, 11:21:28 PM10/15/05
to

The leg of the "T" is split down the middle, then glued back together.
Since splitting follows the grain, and no material is lost to a kerf,
the split is nearly invisible after gluing and sanding.
--
Email reply: please remove one letter from each side of "@"
Spammers are Scammers. Exterminate them.

Nate Perkins

unread,
Oct 16, 2005, 1:09:27 AM10/16/05
to
fredf...@spamcop.net wrote in news:1129331067.123010.77120
@z14g2000cwz.googlegroups.com:

I used to carve the ball and cage that way when I was a kid. I remember
being about 10, and having my grandfather teach me how to do it. We carved
a variety of little things like that. That's more than 30 years ago, and
still one of my best memories of grandpa.

Bruce Barnett

unread,
Oct 16, 2005, 7:33:37 AM10/16/05
to
Doug Warner <dwar...@ccharter.net> writes:

> The leg of the "T" is split down the middle, then glued back together.
> Since splitting follows the grain, and no material is lost to a kerf,
> the split is nearly invisible after gluing and sanding.

Bzzzt. The eBay page says:

"Only two pieces of timber are used, and there are no hidden splits."

bri...@all.costs

unread,
Oct 17, 2005, 4:03:28 PM10/17/05
to
On Sat, 15 Oct 2005 23:21:28 -0400, Doug Warner
<dwar...@ccharter.net> wrote:

>
>The leg of the "T" is split down the middle, then glued back together.
>Since splitting follows the grain, and no material is lost to a kerf,
>the split is nearly invisible after gluing and sanding.


the guy says no hidden splits.

Tim Douglass

unread,
Oct 17, 2005, 6:15:43 PM10/17/05
to

Somewhere I have a bunch of ball-in-cages and chains that I carved
sitting in various classes in High School. Now they would haul me out
and shoot me for having a knife in class. I didn't learn much of the
subject matter, but I got pretty good at carving!

--
"We need to make a sacrifice to the gods, find me a young virgin... oh, and bring something to kill"

Tim Douglass

http://www.DouglassClan.com

0 new messages