# AP-ART

## Stewart T. Coffin

First release of revised edition, December 2000

[Puzzle World Home] [Introduction] [Part 1] [Part 2] [Part 3] [Puzzle Index] [Color Plates]

# Part 2

#### Complete List of AP-ART Designs

Before starting this listing in numerical order, for the sake of completeness I will insert a brief mention of those which predate the list or have been omitted for various reasons that should in most cases be obvious.

The first puzzles that I made were of the common jigsaw variety, when I was barely old enough to dissect scraps of plywood with a coping saw. From earliest childhood, I had a fascination for all things mechanical. My parents were very tolerant of my urge to take apart old appliances and machinery (later radios and electronic equipment), and soon I acquired a knack at repairing them. My father was an early pioneer in both scientific plant photography and pictorial nature photography, but he soon realized the futility of encouraging me to follow in his accomplished footsteps. He plied me with books and magazines on all the things I loved - mechanical, scientific, and mathematical. There was never the slightest question but that I would study engineering in college.

As a carry-over from model airplane days, I had some solid blocks of balsa wood. With these, I was inspired to create some three-dimensional jigsaw puzzles. These probably date from around 1946, since you could not get much balsa wood during World War II.

In that wonderful 1950 book already mentioned, Mathematical Snapshots by Steinhaus, I remember being especially fascinated by the rhombic dodecahedron and its various spatial properties. This must have remained dormant in the back of my mind until reemerging two decades later. That book also described a simple 3x3x3 cubic dissection called Mikusinski's Cube, which sparked my interest in cubic dissections.

The first original design for which I have any record is a 7-piece 4x4x4 cubic dissection called Seven Block. It was designed around 1958 while I was working at M. I. T. Lincoln Laboratory. I include it so that you can see progress has been made since then. Several of us at the Lab were interested in mathematical recreations, especially fellow electrical engineer Gus O'Brien. (It was he who first introduced me to the simple but intriguing six-piece first stellation of the rhombic dodecahedron.) I created the Seven Block and made one for Gus to puzzle over, but he quickly solved it. I made only that one. It is now in a collection in England.

As previously explained, for a while before the woodcraft began, I tinkered with experimental models cast in epoxy. The following were all created between 1968 and 1970.

One of the first was Spinner, consisting of 6 identically shaped pieces in 3 colors, 2 of each. An injection-molded styrene version of this was later produced by Skor-Mor in their Geo-Logic series as Tauri. There was also a 4-color version in which each piece was of 2 halves of different color bonded together, to be assembled in color symmetry. Only about a dozen Spinners were cast.

The  consisted of 12 nearly identical Z-shaped pieces, likewise cast in multicolor, which assembled to form a truncated rhombic dodecahedron. Only a few of this uninteresting design were cast.

Prism consisted of 6 identical pieces, cast in 3 colors, which assembled to form 3 intersecting square prisms. (This was later the basis for the Seven Woods (#42). Only a few of this mundane design were cast.

Pluto was a slightly more interesting version of Prism, in which each piece had a shoulder at one end, with assembled end faces slightly octagonal rather than square, and only one axis of assembly. Only a few were cast.

Octo was similar to Prism except that each piece was dissected longitudinally, making 12 pieces. It likewise used 4 colors with associated color symmetry problems. It was an exercise in dexterity to assemble. One version had a split piece for easier assembly. The assembled shape suggested an octahedron. Only a few were cast, but a modified version later led to the baffling  Three Pairs (#27).

Four Color Cube consisted of 12 cast pieces, 3 of each color, which were to be assembled into a cube with 4 colors on each face. There was also a slightly more interesting version in which the pieces were joined in pairs to make 6 bicolored pieces. Only a few were cast.

Four Color Octahedron was similar to the Four Color Cube described on the previous page except octahedral when assembled, with likewise 12-piece and six-piece versions. Only a few were cast.

Tetrahedron was similar in principle to Prism except that the assembled shape was tetrahedral. It later became the basis for the injection-molded Cetus, produced by Skor-Mor in their Geo-Logic series.

There were many other experimental few-of-a-kind models cast during this phase and on into 1971, mostly not recorded. Some of them became the models for other "puzzles" in the Skor-Mor Geo-Logic series, such as Nova, Spirus, and Uni. There was one, Double Star, that could be assembled inside-out to form two different geometrical solids, but the Skor-Mor version was so misshapen as to be nearly useless.

One of the problems with the Skor-Mor series was that, in order to economize on mold costs, the manufacturer insisted on only puzzles with six identical pieces. In that way, a multi-cavity mold could produce several different puzzles at the same time. There is only so much you can do with six identical pieces. Another problem was that the pieces had to be "cored out," meaning that they consisted entirely of thin walls in order to cool faster in the mold and speed up the mold cycle. I wish now that they had never been made.

The Hectix puzzle, on the other hand, was solid .75-inch hexagonal. This was achieved by including a blowing agent in the styrene and a slower mold cycle, increasing the cost. A lot of this technology was new to the molder, and there were problems with burned or misshapen pieces, some of which were never completely corrected. The discoloration from heat was especially bad because of the ugly off-white color that the manufacturer chose. Our Agreement called for it to be made in four contrasting colors - red, yellow, green, and blue - but they never did it that way. I wish now that it had never been made in plastic. A wooden version would have been better.

As I mentioned in Part 1, for a while in 1970 I even experimented with setting up a production line myself casting puzzle pieces in epoxy. In looking back now, that seems so totally impractical that I can only wonder what I could have been thinking. One problem never overcome was that the RTV rubber molds, which were a lot of work to make, deteriorated rapidly after only a few cycles. The whole process was slow and messy, and most important, the finished product was not very attractive. And so I switched to wood, and the remainder of this publication is devoted to that phase of my work.

Again simply for the sake of being complete, before proceeding to the numerical listing, I will briefly describe just a few of the early attempts in wood which never made it into production:

OCC-Wood consists of 3 plywood pieces in the shapes of O-C-C, plus 2 small cubic blocks. They assemble easily into a familiar burr shape with the blocks inside, and the problem is to have the blocks fall into position to permit disassembly. Without the cubic blocks, it is a familiar old novelty published in many books. Only a few were made around 1973.

Rec-Tangle also consists of 3 pieces in the O-C-C shape as above, but this time of glued up pieces, with a hole in one piece and a dowel loose inside. The first step of disassembly is to shake the dowel into the hole, which can be made as easy or hard as one wishes by the size and shape of the dowel and hole. Four more steps are required to disassemble. There is also a modified version in which the O piece is split into 2 burr-like pieces. Only a few were made around 1973.

Wunder Bar consists of 6 pieces which fit together to form a cubic lattice. There are 4 types of pieces - W, X, Y, Y, Z, Z. Each piece is made up of three 1x1x5 sticks joined together. W and X are mirror image, likewise Y and Z. The diagram shows the design. There are 4 distinct mechanical solutions, but by using multi-colored woods and requiring color symmetry, the number of solutions can be reduced. Only a few were made in 1973.

Interlocking Checkerboard consists of 8 pieces which fit together 2 different ways to form (guess what!) an interlocking checkerboard (1973). Probably none were made.

Cube Brute consists of 24 identical burr pieces which interlock to form a symmetrical cubic-shaped assembly. A 16-piece square assembly is also possible. A couple sets were made in 1973. Pentangle came up with this independently around the same time, sold as their Woodchuck Puzzle. I also proposed a set to be called TWIS-T, in which 16 pieces are bonded together into 8 T-shaped pairs. In yet another variation, RUFTY, the remaining 8 pieces are also bonded together in-line. Multicolored pieces with color symmetry solutions were also proposed. Possibly a few of each were made, all in 1973.

Mosaic is an 8-piece dissection of a 4x4x4 cube, designed in 1979. I must have made at least one, but have no record. The one known solution is shown here, but perhaps there are others. It is almost serially interlocking. An improved version became Convolution (#30).

In the same vein as the above, here is one with no name which appears in my notes for 1979. It is a six-piece 4x4x3 dissection with the capability of being made with eighteen 1x1x2 blocks and twelve 1x1x1 blocks of colorful woods such that symmetrical patterns appear on all faces. It is non-interlocking, so a box or tray might be used to retain it. Must have made one.

Triful was designed in 1973 for production in plastic, and a few models made of colored wood, but it was never produced. It consisted essentially of 12 triangular sticks with end blocks added, in 4 colors, 3 of each. Four pieces, one of each color, were split in two to permit assembly. Then a wooden version was designed around 1975 which used 4 sliding key pieces instead of split pieces, but only one or two were made. Much later this design was resurrected to become the basis for Isosceles (#101) and Iso-Prism (#101-A). See also Notched Rhombic Sticks in PWPD.

As some of the above entries indicate, I have ransacked my records to include everything in this comprehensive listing of AP-ART, no matter how mundane or obscure. There were many others that were never recorded and are now lost and forgotten. A few more may turn up from time to time, and if so I will add them later on as a supplement.

I am not including in this publication any of my admittedly feeble attempts at inventing games, such as Arc-Tic, and Hebee-Shebee, as they do not fit my concept of AP-ART. For the same reason, I omit topological amusements such as Lamplighter, Liberty Bell, Bottleneck, Sleeper-Stopper, Super Sleeper-Stopper, and Figure Eight, most of which are described in PC'85.

Now on to the numbered designs. I have tried to indicate the year designed, published, or first made, when known. The first illustrated brochure I put out was in 1970, and the last in 1990. I have also indicated the approximate number made, but only up to around 1990, after which the quantities decline to the level of a retirement hobby activity.

When a thumbnail photograph of the puzzle is present, you can click on it to go to the Puzzle World page for that puzzle. Check the Color Plates for other photographs.

#### Serial Numbers 1 to 167

1 Ortho-Cube. A 12-piece dissection of a 5x5x5 cube that is not quite solid, 3 kinds of pieces, 4 of each. Symmetrical and not difficult. This early version, which appeared on my first brochure in 1970, was made of .875-inch square birch stock. About 20 made in 1970
1-A The Cube. Same as Ortho-Cube (#1) but made of .75-inch stock in contrasting fancy woods. By using 1x1x3 blocks, symmetrical colorful patterns appear on all faces. Appeared on 1971 brochure. About 100 made, 1971-1972.

This design was later made by Pentangle as their Wookey Hole.

2 Pentablock. The familiar set of 12 solid pentominoes in a 3x4x5 box. Not my design. This one made of .875-inch birch stock. Appeared on 1970 brochure. About 20 made. See PWPD and PC'85.
2-A Pentablock. Slightly improved version of Pentablock (#2), made of .75-inch hardwood stock in a Plexiglas box. Appeared on 1975 brochure. About 10 made.
2-B Pentacube. An improved version of the Pentablock (#2-A), same size, but of 12 contrasting fancy woods colorfully combined in the given solution, with a box of blue mahoe. Appeared on 1977 brochure. About 20 made.
3 Snowflake. Ten flat pieces made of 37 hexagons joined different ways assemble on hexagonal tray to form Snowflake or Hexagon. Other problem shapes shown in booklet. The first version, which appeared on 1970 brochure, was of cast thermosetting resin. About 50 made. See PC'85, PWPD, and PC'92.

The next version was one made by Span-Atwater of cast polyester resin. They made about 500, 1972-1973. Came with 10-page booklet with 49 problems, as did those which follow.

A wooden version is mentioned on my 1977 brochure, but I don't think many were made. The next version was of cast Hydrastone pieces, using surplus Span-Atwater molded styrene bases. Appeared on 1978, 1979, 1981, 1984, and 1985 brochures. About 100 made.

The next version was in fancy wood with plywood base and cover. About 10 made in 1986.

The next version was in thin birch plywood cut by Jim Ayer with a water jet, with molded styrene base. Appeared on 1990 brochure. About 30 made.

The next version was in plastic foam, made and sold by Binary Arts around 1993.

Note: All of these versions except the last one were the same scale - .75-inch hexagons.

4 Sirius (aka Diagonal Star). The familiar six-piece first stellation of the rhombic dodecahedron. Not my idea - in the public domain. See PC'85, PWPD, and PC'92. This version made of glued up pieces in 3 contrasting hardwoods, 1-inch stock. Appeared on 1971 brochure. About 100 made.
4-A Star. An improved version of the Sirius (#4) in 3 fancy woods, 1.25-inch stock. Appeared on 1974 and 1975 brochures. About 400 made.
5 Spider-Slider. Six identically shaped symmetrical pieces, each one made of 4 triangular sticks. In basswood stained 4 colors. About 20 made in 1970, my first polyhedral AP-ART in wood. See PC'85 and PWPD.
5 Scorpius. An improved version of the Spider-Slider (#5) in 4 contrasting fancy woods. Appeared on 1971 and 1972 brochures. About 200 made.
6 Four Corners. Six identically shaped symmetrical pieces in 4 contrasting fancy woods of 1-inch stock, assemble easily to make a polyhedral solid intermediate between the first and second stellations of the rhombic dodecahedron. Appeared on 1971, 1972, 1974, and 1975 brochures. About 200 made. See PC'85, PWPD, and PC'92. A version of Four Corners in 2 fancy woods was made by Roy Rice for the IPP-14 puzzle exchange.
6-A Aries. The plastic version of Four Corners (#6) made by Skor-Mor for their Geo-Logic series.
7 Jupiter. Twelve identically shaped pieces, each one made up of 5 triangular sticks, assemble to form a hollow nearly spherical solid which I call a castellated triacontahedron. Six contrasting fancy woods are used, to be assembled with color symmetry. Over the years, by far my most popular AP-ART creation, but probably as a sculptural curiosity rather than a mechanical amusement. I'll bet half of those made have never been disassembled! Appeared on 1971, 1972, 1974, 1975, and 1977 brochures. About 400 made. See PC'85 and PWPD.
8 Nova. Six identical symmetrical pieces easily assemble to form the second stellation of the rhombic dodecahedron. More of a sculpture than a puzzle. Appeared on 1972 brochure. About 100 made of 1-inch zebrawood.
8-A Nova. A  plastic version of the above made by Skor-Mor for their Geo-Logic line.
8-B Nova. A fancy version in 4 contrasting woods, color symmetry. Three made in 1987.
9 Square Knot. (aka Altekruse). Twelve identical notched square sticks - a popular design patented by W. Altekruse in 1890 and now in the public domain. My version used 3 contrasting fancy woods of .875-inch stock. I also included 2 extra pieces to make my newly discovered 14-piece version. Appeared on 1974 and 1975 brochures. About 40 made. See PC'85 and PWPD.
9-A Frantix. A variation on Square Knot (#9) with pins and holes in place of notches. Two kinds of pieces, 6 of each. Only 4 made in 1973, as it was a prototype for the plastic version, Frantix (#9-B). See PC'85, PWPD, and PC'92.
9-B Frantix. This design was licensed to 3M Company in 1974 for manufacture in injection molded styrene as a sequel to Hectix. It was very poorly made, with tapered holes, and not a success.
9-C Frantix. This more interesting wooden version had extra pins and holes in the centers, 4 kinds of pieces, 3 of each. Probably about 4 made around 1973. See PC'85 and PWPD.
10 Giant Steps. A variation of Square Knot (#9) (but not an improvement!) made by adding extra blocks to 6 of the pieces to make T-shaped pieces. Appeared on 1974 and 1975 brochures. About 20 made, mostly butternut.
11 Hexagonal Prism. Six dissimilar pieces assemble one way only with only one sliding axis - a significant departure from previous AP-ART designs. Appeared on 1974 and 1975 brochures. About 60 made. See PC'85 and PWPD.
11-A Double Hexagonal  Prism. A variation of the Hexagonal Prism (#11) with eight faces made by adding more blocks.
12 Triangular Prism. A variation of the Hexagonal Prism (#11) by adding 12 more blocks to transform it into a most intriguing solid - and one that reappears later as Burr Muda (#112). Appeared on 1974, 1975, 1980, and 1981 brochures. About 100 made, mostly in mahogany. See PC'85 and PWPD.
12-A Triangular Prism (elongate version). This was a slightly different version of the Triangular Prism (#12). Many such variations are possible depending upon how the additional blocks are placed. Only about 2 made around 1974.
13 The General (Four Star). A variation of Triangular Prism (#12) made by adding yet 12 more blocks. Appeared on 1974 and 1975 brochures. About 20 made, mostly in mahogany. See PC'85 and PWPD.
13-A The General (elongate version). Same idea as Triangular Prism (#12-A). Only one made, around 1974.
13-B Ring of Diamonds. An interesting variation of The General ( #13) made using rhombic rather than triangular stick segments. Easier to make. Evidently designed in 1973 and forgotten, then rediscovered in 1995. A few made in 1995. More information on the Dec. 1995 instruction sheet.
14 Super Nova. Same shape as Nova (#8), but 6 dissimilar pieces, difficult. A few had 8 dissimilar woods, but most were one wood. Appeared on 1974 and 1975 brochures. About 20 made.
14-A Second Stellation. An improved version of the Super Nova (#14), more accurately made. Appeared on 1981 brochure. About 50 made in mahogany. See PC'85 and PWPD.

A further improved version of Second Stellation was made starting in 1983, using triangular rather than square stock. About 50 made in mahogany.

14-B Augmented Second Stellation. A variation of Second Stellation (#14) in which the arms are lengthened to make a different shape. Only 2 made in 1990. This was redesigned in 1996 to use smaller .8-inch stock in 4 contrasting fancy woods, and a few more made.
15 Triumph. The 6 identically shaped pieces somewhat resemble those of Four Corners (#6), but have bilateral rather than axial symmetry. Assembles into 3 different symmetrical polyhedral shapes, the first AP-ART to do so. Use of 2 contrasting woods introduces color symmetry patterns also. Appeared on 1974 and 1975 brochures. About 50 made. See PC'85, PWPD, and PC'92.
15-A Fusion-Confusion. By making a Triumph (#15) and bonding two pairs of pieces together different ways, an entirely new and more confusing amusement emerges. It likewise makes the 3 different shapes of Triumph (#15), but the color scheme is different. First made in 1990, and about 40 made in fancy woods. See PC'92.
16 Dislocated Scorpius. A variation of Scorpius (#5) with 6 identical but non-symmetrical pieces, making it slightly more confusing to assemble and disassemble. Appeared on 1974 and 1975 brochures. About 20 made. See PC'85 and PWPD.
17 Dislocated Jupiter. A variation of Jupiter (#7) analogous to Dislocated Scorpius (#16). Most and perhaps all were made of one wood. Appeared on 1974 and 1975 brochures. About 10 made. See PC'85 and PWPD.
18 Abbie's Waffle. Six pieces, each made of 4 cubic blocks, assemble various ways onto a square tray or into a 2x3x4 box. Created by my daughter and demonstrated by her on the PBS children's program ZOOM, first aired 9 Dec 1973. Appeared on 1975 brochure. About 10 made. See PC'85 and PC'92.
18-A Joined Pairs. In the same style as the Abbie's Waffle (#18), 6 pieces, each made of 1x1x2 blocks joined different ways, pack into a 2x3x4 box various ways. Only one made in 1990. See PWPD and PC'92.
19 Pyracube. Four of the pieces are made of truncated rhombic dodecahedron (or edge-beveled cubes) joined different ways, and the 5th piece is a single such block, for a total of 14 blocks. They pack snugly into a their cubic box, with or without the single block, form a square pyramid, a rectangular pyramid, or a triangular pyramid with only 4 pieces. Appeared on 1975 brochure. About 20 made. See PWPD.

A variation of this design using spheres appears in Creative Puzzles of the World by van Delft & Botermans, page 85.

20 Pin-Hole. Six 1x1x3 bars with pins and holes assemble easily into a burr-like figure. By adding one or more pieces twice as long, more complicated assemblies are possible. Appeared on 1977 and 1978 brochures. About 50 sets made. See PC'85, PWPD, and PC'92.
20-A King Pin. This was a variation of Pin-Hole (#20) but with blind holes and free pins as a sort of puzzling large construction set. It never got beyond the development stage in 1975.
20-B Goose. A variation of King Pin (#20-A) but with animated figures, likewise never went beyond the development stage in 1986
21 Cuckoo Nest. Six hexagonal bars are held together with 6 dowels. Two solutions. Appeared on 1977, 1978, and 1979 brochures. About 100 made in birch. See PC'85 and PWPD.
22 Locked Nest. Twelve hexagonal bars are held together with 12 dowels in a symmetrical assembly. There are 2 versions. Most of those made had 5 elbow pieces, but the more challenging version has 6 elbows. Appeared on 1977, 1978, and 1979 brochures. About 100 made in birch. See PC'85 and PWPD.
22-A Locked Nest (three-hole variation). This is simply a version of the Locked Nest (#22) in which the bars and dowels are shortened so that the end holes are eliminated. Only one made in 1990. Never illustrated, but use your imagination.
22-B Siamese Locked Nests. Two Locked Nests (#22) joined together using longer bars. See PWPD. Two made in 1989. (Many other variations are possible.)
23 Scrambled Scorpius. A variation of Scorpius (#6) in which all the pieces are dissimilar and non-symmetrical. Only one solution and one order of assembly. Much more difficult than I realized when I first offered it. One of my most satisfactory AP-ART designs. Appeared on 1978, 1979, 1980, and 1981 brochures. About 200 made. See PC'85 and PWPD.
23-A Egyptian. An improved and enlarged version of Scrambled Scorpius (#23), with the solution coded on the pieces. A few made in oak, 1993-1995.
24 Saturn. Similar to Jupiter (#7) except 6 kinds of pieces, two of each. Intended to have only one solution, but other solutions have been discovered. Appeared on 1978, 1979, 1980, and 1981 brochures. About 65 made, mostly in one kind of fancy wood. The deluxe versions had 6 kinds of wood and doweled joints. See PC'85 and PWPD for description.
25 Hectix. Twelve notched hexagonal bars interlock to form a symmetrical assembly. Nine bars have 2 notches and 3 bars have 3 notches. There are 3 distinctly different solutions. Patent 3721448. One of my most satisfactory AP-ART designs, discovered independently by Bill Cutler around 1965. This is the plastic version licensed to 3M Company in 1970. About 100,000 made. It was supposed to be in 4 colors, but they never got it. See PC'85 and PWPD. Copies have been made in Japan, France, and Australia.
25-A Hexsticks. This is my wooden version of the Hectix (#25)in which 7 bars have 2 notches, 3 bars have 3 notches, and 2 bars have one notch. Same solutions possible. Appeared on 1979 and 1981 brochures. About 150 made, mostly birch.
25-B Giant Hectix. Wooden version of the original Hectix (#25) but twice the size - 1.5-inch hexagonal stock. A few made in 1993.
25-C Four-Color Hexsticks. A wooden version of the original Hectix (#25), but in 4 contrasting colors as it was supposed to have been, and larger - 1-inch instead of .75-inch. A few made in 1995.
25-D Hextix. A wooden version made by Bits & Pieces and seen by me only in their 1996 catalog.
26 Four-Piece Pyramid. Four pieces, each made of 4 rhombic dodecahedron blocks joined different ways, assemble with surprising confusion one way and one order only to build a pyramid. Appeared on 1979 and 1981 brochures. The first 12, made in 1976, used edge-beveled rosewood cubes and doweled joints. The next 30, made around 1979, used 1-inch cherry rhombic dodecahedron blocks. About 25 were made starting in 1981 with larger edge-beveled cubes. In 1997, I made a few in contrasting fancy woods. See PC'85 and PWPD.
27 Three Pairs. Two kinds of pieces, 3 of each, assemble with amazing difficulty, to make the solid shown. My first truly coordinate motion puzzle, and still one of the best. Appeared on 1979 and 1981 brochures. About 150 made in mahogany. In 1986, a deluxe edition of 10 were made in rosewood with doweled joints. See PC'85, PWPD, and PC'92.
27-A Three Pairs Variation. (aka Split Second) Several variations are possible, all with the same internal function but different external shape. The one with this designation is the second stellation of the rhombic dodecahedron, only one model made. I have also made a model having the shape of the first stellation of the rhombic dodecahedron, and another resembling the Four Corners (#6). Used as an exchange puzzle at IPP-19.
28 Truncated Octahedra. Five pieces made of 14 truncated octahedra blocks pack into a square box. Twelve-page booklet shows 18 other problems. Appeared on 1979 and 1981 brochures. About 50 made using 1.5-inch mahogany cubes with corners removed and box of Baltic birch which could be inverted to hold the square pyramid construction. See PC'85 and PWPD.
29 Half-Hour. Six-piece 3x3x3 cubic dissection with unique solution. Dozens of other construction problems were submitted by H. Havermann and D. Barge. Appeared on 1980 and 1981 brochures. About 50 made of 1-inch stock. Reissued in 1984 using 6 contrasting woods and a box of blue mahoe. See PC'85, PWPD, and PC'92.
30 Convolution. Seven pieces assemble one way only and in one order only to form a 4x4x4 interlocking cube with symmetrical pattern on all 6 faces. Appeared on 1980 and 1981 brochures. About 50 made of .75-inch stock. See PC'85, PWPD, and PC'92. Recently a fine reproduction of this design has been made by Wayne Daniel.
31 Octahedral Cluster. Four pieces made of 19 rhombic dodecahedron blocks assemble one way only and in one order only to form an interlocking octahedral cluster. Baffling. Appeared on 1980 and 1981 brochures. About 40 made of 1.25-inch limba rhombic dodecahedron blocks. See PC'85, PWPD, and PC'92.
31-A Five-Piece Octahedral Cluster. Like the Octahedral Cluster (#31) except 5 pieces. Perhaps even more baffling. Recently I have made a few from 1-inch edge-beveled cubes of 5 fancy woods.
32 Broken Sticks. Six dissimilar pieces made of triangular stick segments assemble one way only with one sliding axis. Difficult. Appeared on 1980 and 1981 brochures. About 50 made. See PC'85 and PWPD.
33 Twelve-Point. Six dissimilar pieces assemble one way only to form an intriguing solid intermediate between 2nd and 3rd stellations of the rhombic dodecahedron. Appeared on 1981 brochure. About 50 made of 1-inch stock in 2 contrasting woods. See PC'85 and PWPD.
33-A Twelve-Point. A remake of the Twelve-Point (#33) in .8-inch stock. A couple made in 1996.
34 Augmented Four Corners. Six pieces assemble one way only with one sliding axis. Appeared on 1981 brochure. About 60 made of 1-inch stock in 2 contrasting woods. See PC'85 and PWPD.
34-A Modified Augmented Four Corners. These are variations of the Augmented Four Corners (#34) made by sanding the four faces down by varying amounts to create interesting sculptural effects, especially when using contrasting woods. Two versions made, one model of each, probably in early 1970's. One is shown.
35 Burr #305. Ordinary notchable Six-Piece Burr with unusual 3+3 solution. Appeared on 1981 brochure. About 60 made of 1-inch stock. See PC'85, PWPD, and PC'92.
36 Coffin's Improved Burr. More complicated than Burr #305 (#35), with 5 shifts to release the first two pieces. Appeared on 1981 brochure. About 50 made. See PC'85.
37 Star-of-David. Six dissimilar non-symmetrical pieces assemble to form three different interlocking polyhedral solids with confusing diagonal axes of assembly. Appeared on 1981 brochure. About 50 made of 1-inch mahogany stock. See PWPD.
37-A Star-of-David, improved version. Same as Star-of-David (#37) but achieved with simpler pieces. About 12 made, beginning in 1990. In 1997, a few made of .8-inch stock in 2 contrasting fancy woods. See PC'92.
38 Three-Piece Block. Three pieces made of 10 cubic blocks interlock to form a triangular pyramid. Surprisingly confusing. First 300 made of cherry for Citibank in 1980. About 50 more made starting in 1981, all of 1-inch blocks. See PC'85, PWPD, and PC'92.
39 Rosebud. Two kinds of pieces, 3 of each, assemble with much difficulty by coordinate motion to form an intriguing polyhedral solid. About 42 made, starting in 1982. Some were in mulberry and cherry, others the deluxe in rosewood and tulipwood. See PWPD.
39-A Rosebud Assembly Jig. Was made available later. Could (perhaps) be assembled without, but much easier with. About 32 made, starting in 1983.
40 Interrupted Slide. A higher-level Six-Piece Burr with one or two interesting solutions, depending upon the length of pieces. Made 28 in 1982 of golden bilinga. See PC'85.
41 Unhappy Childhood. Ten checkered pieces, each made of 5 cubic blocks joined different ways, pack into a 5x5x2 box. About 50 made, 1983-1984. See PWPD, page 47.
42 Seven Woods. Six simple pieces make a modified Diagonal Burr. About 20 made in 1971. See PC'85
42-A Brickyard. A Seven Woods distorted by flattening along vertical axis. Only 2 made. Proposed for 1996 exchange puzzle but never used. Too simple perhaps.
43 Sleeper-Stopper. (Topological - not AP-ART)
44 Super Sleeper-Stopper. (Topological - not AP-ART)
45 Buttonhole. (Topological - not AP-ART)
46 Vega. Six simple identical pieces assemble easily to form an attractive solid. About 30 in fancy woods, 1-inch stock, sold at craft shows, 1972-1975.
46-A Vega II. Same shape as Vega (#46), but made by truncating Superstar (#50). Maybe made one or two.
47 Cluster-Buster. Six identical pieces assemble easily into a polyhedral solid, but trickier to disassemble. About 5 made in 1973. See PC'85 and PWPD, page 97.
48 Truncated Cluster-Buster. Minor variation of the Cluster-Buster (#47). About 5 made in 1973.
49 Improved Cluster-Buster. Variation of the Cluster-Buster (#47) with dissimilar pairs of pieces. An unpublished design. About 10 made in 1973.
50 Superstar. Six identical pieces assemble easily to make 3rd stellation of rhombic dodecahedron. About 10 made in mahogany for craft shows, 1972-1975. See PWPD.
50-A Superstar II. Variation of the Superstar (#50). One made in 1990. See PWPD.
50-B Third Stellation. Version of Superstar II (#50-A) in 4 contrasting fancy woods, to be assembled with color symmetry. Proposed in 1986 but never made. See PWPD.
51 Little Superstar. Truncated version of Superstar (#50). Six identical pieces make 2nd stellation of the rhombic dodecahedron. Just a sculpture. Probably made a few.
52 Pennyhedron. Two pieces form rhombic dodecahedron. Amusing to take apart. Invented by my kids in 1971 while playing with scraps. With variations too numerous to mention. Made about 150, 1971-1985. See PC'85, PWPD, and PC'92.
52-A Hole-in-One. Simple 3-piece derivative of Pennyhedron (#52) with pin and hole, coordinate motion. Proposed exchange puzzle but never used. Unpublished design.
52-B Button Box. A distorted Pennyhedron (#52)in 4 fancy woods, used as an exchange puzzle at IPP-16.
52-C Pennyhedron Trick Pair. A pair of Pennyhedrons which look alike but come apart two different ways. Only one set made. See PC'85 and PWPD.
53 Little Giant Steps. An uninteresting variation of Giant Steps (#10) in which the added blocks are cubic instead of 1x1x2. About 3 made in 1973.
54 Defiant Giant. A bizarre variation of Giant Steps (#10) in which the added blocks are attached differently. The first piece is a standard Square Knot (#9) piece of length 5, of which six are required. The next piece has a 1 x 1 x 1.5 block attached, and the next two a 1 x 1 x 2 block, one of each required. The last piece also has a 1 x 1 x 2 block, three required. Very tricky to assemble. Not previously published. Only one made in 1973.
55 Pagoda. Eight cubic blocks are added to the Square Knot (#9), resulting in 3 kinds of pieces, 4 of each. Somewhat trickier than standard Square Knot to assemble. Probably a few made in 1973.
56 Giant Pagoda. This version combines both the added blocks of Giant Steps (#10) and Pagoda (#55), resulting in 6 kinds of pieces, 2 of each. Probably made a few in 1973.

One of the delights of this editorial project is rediscovering old models, or plans for same, that were filed away many years ago and forgotten. Usually the reasons for putting them aside are obvious. On the other hand, in the light of experience sometimes new ways are seen to correct some deficiency. Every design listed here has its roots in previous designs, sometimes recombining in unexpected ways. Alas, such is not the case with the four dull ideas listed above, at least not yet.

57 Plus 2. The 14-piece version of Square Knot (#9). See PC'85 for the amusing story of this design and its many variations. I may have made about 20 of these, 1973-1975, but my records did not always distinguish between the 12- and 14-piece versions. I made a few of the larger versions also, such as the 24- and 36-piece.
58 Diagonal Cube. Six pieces in 2 contrasting woods assemble diagonally to form an attractive cube. About 20 made in fancy woods, 1981-1985. See PC'85 and PWPD.
59 Corner Block. A Pin-Hole (#20) with 8 cubic blocks added to the corners. About 30 made in, starting in 1980. See PC'85.
59-A Improved Corner Block. Designed in 1985 to replace Corner Block (#59). About 15 made. See PC'92.

A nice replica of Corner Block (#59) has been made by Wayne Daniel

60 Garnet. Six dissimilar pieces, each made of 4 tapered blocks, assemble into a rhombic dodecahedron. Appeared on 1984 brochure. About 30 made. See PWPD.
61 Setting Hen. Four pieces made of 14 rhombic dodecahedron blocks fit into a cubic box and construct other problem shapes. Appeared on 1984 brochure. About 30 made. Design not published.
61-A Distorted Cube. A revised and improved version of the Setting Hen (#61), made instead with edge-beveled cubic blocks. Tricky box makes both cubic or rectangular insides. Designed in 1988. About 20 made. See PC'92.
62 Nine Bars. Same idea as Cuckoo Nest (#21), but 9 bars and dowels instead of 6, and 3 layers instead of 2. Only one solution known - difficult. Appeared on 1984 brochure. About 10 made in birch. See PWPD.
63 Pseudo-Notched Sticks. Six identical pieces. A novelty. Looks like ordinary Diagonal Burr but comes apart differently. About 25 made in 1985. See PC'85 and PWPD.
64 Expanding Box. Just a novelty. Six identical pieces enclosing cubic space expand with coordinate motion. Two made in 1971. See PWPD.
65 Thirty Notched Sticks. In my file of design ideas for 1972, I list theoretically a rhombic version and a couple of pentagonal versions. Shown are a cast epoxy triangular version and a rough wood pentagonal  version. For more information, see PWPD.
65-A Thirty Notched Rhombic Sticks. This version of the Thirty Notched Sticks (#65) has an instruction sheet dated 1987, evidently the year that a few were made. See PWPD.
66 Crystal Blocks. The pieces are rhombic dodecahedron blocks joined together different ways. Originally cast in epoxy, this would be listed in Part 1 except that I tinkered with wooden versions also. There were various sets tried, but the final version had 6 pieces and 22 blocks. It never got beyond design stage, but still it was an entertaining project, I proposing the sets of pieces and their associated problems, and Mike Beeler analyzing them using his computer. Only a few sets were actually made, either cast epoxy or wooden. But who cares if it never went into production. We had fun doing it! I still have all the papers, in case anyone wants to pursue something along these lines.
67 Peanut. Six polyhedral pieces fit together many ways to construct problem shapes. Designed in 1973 for possible manufacture in plastic. Resurrected in 1986 to make in wood. Made about 30 in 1-inch mahogany. See PC'85, and PC'92.
67-A Shatterblock. Same idea as Peanut (#67), but based on a different dissection of the rhombic dodecahedron - three-pronged as in Pennyhedron (#52). Five kinds of pieces, two of each. Designed in 1973, and originally intended for production in plastic, which influenced the choice of design for ease of injection molding. Never got beyond design stage. Unpublished.
67-B Pennydoodle. Improved version of Shatterblock (#67-A). Five pieces fit together many different ways. Eight problem shapes shown. About 30 made, 1989-1990. To be entirely satisfactory, they had to fit together precisely. But they had a tendency to warp and then not fit, so discontinued making. See PC'92.
68 Confessional. Variation of Square Knot (#9) using 85 degree rhombic sticks rather than square, .75 x 3 inches. Two kinds of pieces, 8 of one and 4 of the other. Tricky solution involves rotation. Designed in 1994.
68-A Leaning Tower of Altekruse. A 14-piece variation of Confessional  (#68-B). An IPP-15 exchange puzzle.
68-B Confessional (long version). Like the original Confessional (#68) except sticks are 5 units long instead of 4. Even more confusing. One of my most satisfactory designs, but only a few made because of the labor of sawing out the slant notches with multiple saw cuts. With a special cutter, would be easy.
69 Unnamed. Distorted variations of Scorpius (#5) and Jupiter (#7) that never got beyond preliminary design stage.
70 Improved Saturn. An improved version of Saturn (#24) with unique solution. Never found one. See PWPD. Also see Sphinx (#164)
71 Stucksticks. A  Hexsticks (#25-C) in 4 colors, but with 4 pairs of pieces bonded together for greater difficulty. Designed in 1995, and so far only 3 made. Unpublished.
71-A Stucksticks (harder version). Variation of Stucksticks (#71), with only one kind of wood, even more confusing. Proposed but not yet made.
72 Design Number 72. Based on a dissection of the triacontahedron into 60 identical wedge-shaped blocks. In this version, each piece made of 6 blocks, all pieces dissimilar and non-symmetrical, and each a different kind of wood. Six made in 1985. Unpublished design, but this illustration gives a hint.
72-A Design Number 72 (two-tiered version). Very complicated. Evidently one made in 1990, unpublished and unrecorded. In the two-tiered version, the blocks of the outer shell are truncated to make space for the inner layer. The inner layer are truncated also, because easier to make that way.
73 Seven-Piece Third Stellation. Tricky! All 7 pieces dissimilar and non-symmetrical. First step is 3-piece coordinate motion, remaining steps serially interlocking. Evidently designed in early 1970's and forgotten. Rediscovered in 1985, and five made in 1-inch mahogany. A satisfactory but unpublished design.
73-A Seven-Piece Third Stellation (modified). This version of the Seven-Piece Third Stellation (#73) was designed in 1996 to incorporate a very tricky rotation in the coordinate motion first step of assembly. Unfortunately, this results in 2 pieces being identical and the final piece being a plain straight stick. A few made in .8-inch fancy woods, four colors. Unpublished design. I think I prefer the original version.
74 Square Face. To the Pseudo-Notched Sticks (#63), 12 additional notched blocks are attached, making 6 dissimilar non-symmetrical pieces. Two solutions. About 20 made, 1987-1988. See PWPD.
74-A Square Face (improved). A variation of Square Face (#74), likewise 6 dissimilar non-symmetrical pieces, one solution with minor variations, coordinate motion.
75 Split Star. Complicated two-tiered variation of Garnet (#60), with the added outer layer making the first stellation of the rhombic dodecahedron. Four made in 1985 of applewood. See PC'85.
75-A Two Tiers. Could be considered two Garnet (#60) types, one inside the outer hollow one. Designed and illustrated in 1988, but never made. See PWPD.
76 Cornucopia. A large family of designs involving fitting 10 hexominos chosen from a special set of 17 onto an 8 x 8 tray. Analyzed by M. Beeler around 1985 using a powerful computer and sophisticated techniques. About 100, all different, made 1985-1987. Also made in kit form. See PWPD.
76-A Cornucopia 105747 (the Copious Cornucopia). Selected from among the 8203 possible usable sets for greatest versatility of solutions. About 30 made, 1985-1986, using 10 contrasting fancy woods.
76-B Cornucopia 107715. Used in the IPP-16 puzzle exchange.
77 Pieces-of-Eight. Eight dissimilar pieces join together different ways to construct a cube and 8 other shapes. About 25 made, 1986-1988. See PWPD.
78 Pillars of Hercules. Has superficial appearance of an ordinary 5-piece dissection of a 3x3x3 cube, but two of the pieces break in two, using a joint as in Pieces-of-Eight (#77), to make 7 pieces, greatly increasing the difficulty. Two made in 1990. See PWPD.
78-A Pillars of Hercules. Similar to the Pillars of Hercules (#78) except that the three apparent pieces all break in two, making six puzzle pieces. Trickier, since one piece must be assembled last and come apart first. Three made in 1990.
78-B Pillars of Hercules. Another variation of the Pillars of Hercules (#78) in which, instead of coming apart, two of the pieces have rotating joints, made using countersunk flat-head screws, with the screw heads hidden by glue joints. Two made in 1990.
78-C Five-Piece Solid Block Puzzle. Just an ordinary 5-piece dissection of the 3x3x3 cube. What makes it different from most is that if accurately made it tends to be interlocking, depending upon one's definition of interlocking, and surprisingly confusing. Two made in 1990. See PC'92.
78-D Pretty Puzzle. Another 5-piece dissection of the 3x3x3 cube, this one not very difficult but if made of colorful contrasting woods has symmetrical patterns on the faces. One made in 1990.
79 Triple Cross. Among the various possibilities of using larger pieces with the type of joint used in the Pieces-of-Eight, this particular one used 12 or 14 pieces assembled in the familiar Altekruse configuration. Two made in 1973. See PWPD, page 165.
80 Thirty Pinned Pentagonal Sticks. This creation uses 30 identical pentagonal sticks pinned together with 30 dowels into a sort of spherical cluster. Each stick has 7 holes. It could be considered a geometrical construction kit, not too difficult if you have an illustration of the assembly to follow. About 20 made, 1987-1988. See PWPD.
80-A Thirty Pinned Pentagonal Sticks (smaller 5-hole version) The sticks are shortened and the two end holes omitted. One made in 1988. See PWPD.
80-B Thirty Pinned Pentagonal Sticks (even smaller 3-hole version). One made in 1988. See PWPD.
81 Nest Construction Set. This idea was proposed as an extension of the scheme used in the Locked Nest. No sets were ever produced - only a few experimental pieces. See PWPD.

81-A,
81-B-1,
81-C-1.
Two-Three, Four-Legged Stand and Double Four-Legged Stand. These were all minor variations of the Nest Construction Set (#81). See PWPD.
82 Patio Block. Eight pieces, each one consisting of 1x2x2 rectangular blocks joined different ways, assemble to form a 4x4x4 cube with external symmetry. The idea for this came to me in a publication by Ric van Grol, which attributes the original 10-piece design to Toshiaki Betsumiya and an 8-piece version to Kevin Holmes. Four made in 1988. See PWPD and PC'92.
83,
83-A
Pentagonal Stand. Five identical pentagonal sticks are held together with five dowels. In the "A" version, two of the dowels are fixed in place to make two elbow pieces. The assembly has five-fold symmetry. About 20 made in 1990. See PWPD.
84 Obstructed Pins. This could be considered a variation of Locked Nest (#22), with three holes in each of the 12 sticks rather than five, and no elbow pieces. Three of the sticks are slightly shorter at one end, allowing three dowels to be removed to begin disassembly. Two made in 1988.
84-A Unnamed. This is a 30 pentagonal stick version of the Obstructed Pins (#84). Details of the design were never recorded. Only one made in 1988, now in California.
85 Twelve-Piece Separation. Twelve triangular sticks with pyramidal end blocks attached assemble with some difficulty to form an interlocking symmetrical burr. Fully described in PC'92, including solution. About 30 made, 1988-1992. A 1997 version was 10% smaller with fancy woods, 5 made.
85-A Geodynamics. This is a variation of the a Twelve Piece Separation (#85) where it has been distorted by expansion along one orthogonal axis and compression along another - i.e. from cubic to brick shaped. To put it another way, all the sticks are of 50-60-70-degree cross-section. Very confusing to assemble, but the pieces are lettered in conjunction with printed solution. Difficult to make, requiring many complicated saw jigs. Thirteen made in 1994. There was a variation made of the Geodynamics, which my notes indicate had additional blocks for even greater confusion, details not recorded. One made of Sitka spruce around 1990, now in England.

Note: From here on, regrettably many of the descriptions will be found to be lacking in sufficient detail to serve as directions for making them. The reason for this is that we now enter a phase of AP-ART in which many of the designs involve distortions, weird angles, or other complications - hard to design, even harder to make (not to mention solve!), and beyond my patience to attempt to illustrate.