1
00:00:12,640 --> 00:00:19,750
Well, thank you, Sam, for those flattering comments. And I'm sorry you didn't mention about me being late for lectures and things like that.

2
00:00:22,360 --> 00:00:29,350
I want to talk about, well, what's you see on the screen that we've just that that comes a little bit later.

3
00:00:29,830 --> 00:00:36,370
But eventually I shall talk about the pattern which you see in the front of a building, the paving there.

4
00:00:36,520 --> 00:00:39,760
And I want to explain how that came about and what it's all about.

5
00:00:40,450 --> 00:00:49,900
But before getting to that, I want to make some more general comments first about crystal symmetries and then about non crystal symmetries.

6
00:00:50,950 --> 00:01:01,089
So these things are, I assume, very familiar to people that you can have a plain pattern, which is it's has a translational symmetry.

7
00:01:01,090 --> 00:01:05,230
So you can slide this along in some way, and it's the same as it was before.

8
00:01:05,860 --> 00:01:09,010
And in addition to that, has some rotational symmetry.

9
00:01:09,670 --> 00:01:12,400
Now there's a theorem, that very ancient theorem, I think,

10
00:01:12,880 --> 00:01:17,890
which the only rotational symmetries you can have will be two, four, three, four, four, five and six, four.

11
00:01:18,520 --> 00:01:22,510
And just to illustrate those different symmetries, we have them here.

12
00:01:23,620 --> 00:01:29,799
If you have a parallelogram and a pattern of them arranged like this, then about the centre of that,

13
00:01:29,800 --> 00:01:34,900
you're going to have a two fold symmetry rotation through 180 degrees since it's into itself.

14
00:01:36,310 --> 00:01:39,490
Equilateral triangles again, take the centre of one of the triangles.

15
00:01:39,880 --> 00:01:43,250
Then you can rotate the whole pattern into itself 120 degrees.

16
00:01:43,250 --> 00:01:49,810
So it's three times near back to where you start for follow the very familiar square pattern and six fold hexagons.

17
00:01:49,840 --> 00:01:54,400
Now these are all very familiar. Why are they the only ones as well?

18
00:01:55,150 --> 00:02:03,220
There are various ways you can prove that. One of the simplest is as follows Let me just get the stuff working here.

19
00:02:11,030 --> 00:02:17,220
Suppose you have some pattern which has both a rotational symmetry and a translational symmetry.

20
00:02:17,990 --> 00:02:24,920
And I suppose there discrete points which don't sort of crowd in infinitely infinitesimally close to each other, something like that.

21
00:02:24,930 --> 00:02:27,200
So we'd have a, a set of points on the plane.

22
00:02:28,040 --> 00:02:38,089
And I'm going to suppose that I found a point which, about which the pattern has a symmetry, which is one of these symmetry.

23
00:02:38,090 --> 00:02:46,160
So n is the is an info symmetry. So if you rotate through 380, 360 degrees over n pattern, we're going to itself.

24
00:02:46,430 --> 00:02:49,580
So that's a point in the plane with that that property.

25
00:02:50,180 --> 00:02:54,320
And here's another one. And I'm going to choose those as close as possible.

26
00:02:54,740 --> 00:02:59,480
So let's find somewhere where I find two such symmetry points.

27
00:02:59,690 --> 00:03:02,360
If there's one, there's going to be another, because I've got a translational symmetry.

28
00:03:02,360 --> 00:03:07,970
So you move this along somewhere and there'll be another point about which it has rotational symmetry.

29
00:03:08,330 --> 00:03:11,030
And we're going to find two which are as close as they can be.

30
00:03:11,990 --> 00:03:21,410
Now, I rotate this one about that one by the symmetry 360 over N and then that since the pattern is the same when you rotate through that,

31
00:03:21,410 --> 00:03:27,710
that also must be an end fold symmetry point. Then I look at this one and I rotate in the opposite direction.

32
00:03:28,280 --> 00:03:31,340
This one. And so that one must be an info symmetry point.

33
00:03:31,730 --> 00:03:37,130
And you see those are closer together, which contradicts these two being as close as possible.

34
00:03:37,760 --> 00:03:43,470
With the only exceptions being, well, maybe they're not closer together, which would be the case if n equals two.

35
00:03:43,510 --> 00:03:47,450
Then of course, they're not an equal three. They're not either and equals four.

36
00:03:47,800 --> 00:03:50,900
They're exactly the same distance and equal six.

37
00:03:50,900 --> 00:03:55,400
You get away with it because they coincide. And so those are the only possibilities.

38
00:03:56,060 --> 00:04:01,670
If you have an equals five, you can see they're obviously closer here.

39
00:04:02,150 --> 00:04:06,800
If n were larger than six, they'd cross over, but they'd still be closer together, be more like that picture.

40
00:04:07,520 --> 00:04:11,150
Okay, so that's the proof. Really? Very simple. Not much.

41
00:04:12,980 --> 00:04:18,260
You can understand that pretty reasonably, I hope. Now, what about this pattern?

42
00:04:20,420 --> 00:04:28,970
Well, it has a look as though it's both symmetrical in the sense you translate it and it goes into itself.

43
00:04:29,390 --> 00:04:32,440
And five fold rotational symmetry.

44
00:04:32,930 --> 00:04:37,760
All of these stars, pentagons and things like that. In fact, both those statements are almost true.

45
00:04:38,210 --> 00:04:42,230
So this pattern could be extended to infinity. And I'll show you the way you can do that.

46
00:04:42,530 --> 00:04:49,219
Extend it to infinity. And we have the property that if you give me any percentage less than 100%,

47
00:04:49,220 --> 00:04:56,080
so 99.9%, then I would be able to find a way of sliding that pattern along itself,

48
00:04:56,090 --> 00:05:03,720
a translation symmetry, so that it goes into itself to 99.9%, that is to say, of the line segments.

49
00:05:03,740 --> 00:05:11,360
99.9 of them would be in exactly the same place as before, and only 1% would be different.

50
00:05:12,170 --> 00:05:16,130
And you imagine and I want it better than that. How about 99.99%?

51
00:05:16,550 --> 00:05:23,960
So I say, okay, yeah, I can do that too. And I find another translation and it will agree to 99.9% and also I should say rotation.

52
00:05:24,140 --> 00:05:33,770
I can find points about which it will rotate and the symmetry in the first case would have been 99%, 99% secondary, 99.9%.

53
00:05:34,190 --> 00:05:37,399
And you might say, well, I'm no, I'm at 99.999999.

54
00:05:37,400 --> 00:05:42,370
What have you like? There will be a translation and a rotation of this pattern.

55
00:05:42,890 --> 00:05:48,110
Five fold rotation of symmetry and a translational symmetry to that precision.

56
00:05:48,560 --> 00:05:55,070
So it's almost in that sense, never quite exactly 100%, but anything slightly shorter?

57
00:05:55,070 --> 00:06:02,059
100%. Yes, you can find find it. Well, now, you might go back to the original argument and say, well, what goes wrong with the argument?

58
00:06:02,060 --> 00:06:12,650
Let's take it. So we're looking at not exactly symmetric points, but we're looking at points which are has the symmetry to 99.9% to something.

59
00:06:13,280 --> 00:06:18,120
The thing is that if those are 99.9. Percent points.

60
00:06:18,750 --> 00:06:22,680
These ones are likely to lose just a little bit of symmetry because this one's only 90.9.

61
00:06:22,950 --> 00:06:25,980
99.9. That'll probably be 99.8 around.

62
00:06:26,430 --> 00:06:33,450
So you lose just slightly, slightly, they'll get closer. But you'll have lost just a little bit of the symmetry, the precision.

63
00:06:33,450 --> 00:06:37,620
The presentation will be slightly less. And you keep doing that.

64
00:06:37,620 --> 00:06:44,850
And of course, by the time you've got down to the size of the percentages here, you will have lost all all the precision.

65
00:06:45,570 --> 00:06:53,030
So you can see that it slips through the proof as well. Illustration is why when you give a proof of something, you really mean what you say.

66
00:06:53,040 --> 00:06:56,490
It's got to be the statement has to be exactly what you say.

67
00:06:56,610 --> 00:07:00,600
The if it's only APR, you may find that there are loopholes.

68
00:07:00,780 --> 00:07:07,990
And this is a good example of that. Now, let me tell you how this pattern is constructed.

69
00:07:10,030 --> 00:07:11,860
It's very simple in principle.

70
00:07:12,940 --> 00:07:22,850
Let's take the regular Pentagon and what I'm going to do as I'm going to subdivide that Pentagon into six smaller ones, which almost fill it.

71
00:07:23,470 --> 00:07:27,970
So I can find out where this one is by joining those two points, if you like, and that tells you what that line is.

72
00:07:28,390 --> 00:07:36,130
So if I join none adjacent points, I can see where this central Pentagon is, and then I join those and I can find those lines and so on.

73
00:07:36,490 --> 00:07:43,030
So I find that there are six regular pentagons which almost fill that Pentagon.

74
00:07:43,840 --> 00:07:50,830
Now, what I'm going to do is to blow this up to the scale so that this one in the middle is the now the same scale as that one.

75
00:07:51,730 --> 00:07:55,120
And then do it again. But up. Do it again. Okay.

76
00:07:56,530 --> 00:08:00,780
Now, if I. I think.

77
00:08:01,620 --> 00:08:04,980
Yes. Here we go. If I do that here, too, this picture.

78
00:08:07,470 --> 00:08:11,670
Then this Pentagon. That's the big Pentagon. That was that one. And that Pentagon.

79
00:08:11,670 --> 00:08:16,140
There was that one. I mean, that one say.

80
00:08:17,330 --> 00:08:22,850
And then I subdivide these ones and you see this little gap in the middle there, which is a bit of a nuisance.

81
00:08:23,330 --> 00:08:30,950
Never mind. We're almost there. But what I'm going to do now is imagine that we've subdivided again.

82
00:08:31,910 --> 00:08:37,370
And then this one would be subdivided. So there'll be a little triangle pointing in their little triangle pointing out there,

83
00:08:37,370 --> 00:08:42,440
and I'll have a sort of spiky rhombus the next time around, and it'll look like that.

84
00:08:43,100 --> 00:08:49,430
So this is the this big Pentagon there is but almost half the picture here.

85
00:08:49,730 --> 00:08:53,840
But so that one there is now become this one and so on.

86
00:08:54,590 --> 00:09:00,860
And there's my spiky rhombus in the middle. And I find that there's exactly room for another one of these pentagons down there.

87
00:09:01,730 --> 00:09:06,740
I still have two gaps. This well, I call that a pentacle.

88
00:09:07,190 --> 00:09:10,250
If the lines have gone through each others what people call a pentagram.

89
00:09:10,640 --> 00:09:14,150
But this is a pentacle. It goes in and out like that. Okay. It's a star shape.

90
00:09:15,080 --> 00:09:20,300
And the other one will I call it a justice camp? I think that's one way to describe it.

91
00:09:20,930 --> 00:09:24,559
There's the jester, if you like. Okay. Okay.

92
00:09:24,560 --> 00:09:31,250
So we're going to keep doing that, blow it up again the next time the pentacle will get a little spikes like that.

93
00:09:32,030 --> 00:09:41,480
And the nice thing about it is that I can always just find room for pentagons.

94
00:09:42,530 --> 00:09:50,180
You see, if I take the pentacle and I put the spikes on the next the next stage where I subdivide all these, there's little spikes sticking out there.

95
00:09:50,420 --> 00:09:54,620
I have another pentacle in the middle. And three, five more just as caps.

96
00:09:55,370 --> 00:10:00,229
And with the jester's cap here, the other way up in the mind is that just this cap?

97
00:10:00,230 --> 00:10:06,620
There I find three pentagons, one pentacle and three just as caps.

98
00:10:07,370 --> 00:10:12,080
And with the rhombus, one pentagon, one pentacle and one jester's cap.

99
00:10:12,560 --> 00:10:18,140
So I don't have any new shapes. So each time I can blow it up and subdivide the ones I had before.

100
00:10:18,470 --> 00:10:27,590
According to this pattern, there's only one little thing much where you might not, but it's more likely to worry you.

101
00:10:27,590 --> 00:10:31,910
I think if you if you, you know, mathematicians like to worry about things like this when they don't need to.

102
00:10:32,780 --> 00:10:36,470
There's an ambiguity here. You see, this could have been done the other way up.

103
00:10:37,220 --> 00:10:43,850
And you say, well, okay, do it one way. No, I'm going to be I have a rule which makes precise which way I do that.

104
00:10:44,780 --> 00:10:48,860
And what is that rule? Well, I'm going to adopt the following rule.

105
00:10:49,490 --> 00:10:53,330
Looks a bit funny here. I'm just telling you this,

106
00:10:53,330 --> 00:11:00,680
but it's not hard to see that wherever you find one of these spiky rhombus is that Pentagon either on one side or the other.

107
00:11:00,770 --> 00:11:04,429
Rhombus, I should say spiky rhombus, one side or the other.

108
00:11:04,430 --> 00:11:08,570
You will find this pattern and this pattern. You look over there and see where that one is.

109
00:11:08,720 --> 00:11:12,960
If it's not on this side, you'll find on the other side. And then there's a little rhombus in the middle.

110
00:11:13,730 --> 00:11:21,170
And the argument is that whichever way we do, this is governed by where that Pentagon is.

111
00:11:21,200 --> 00:11:30,740
So you take a symmetry about the major transverse, what do you call it, diagonal of this rhombus here.

112
00:11:31,130 --> 00:11:34,430
And this thing flips over a mirror symmetry about that.

113
00:11:34,430 --> 00:11:37,490
And that tells you where that Pentagon goes. So this one is down there.

114
00:11:37,730 --> 00:11:41,900
That one's down. If it were up there, that won't be up. That's the correct rule.

115
00:11:42,440 --> 00:11:46,990
That one's wrong. Now, why didn't I say this is the correct rule and that one's wrong?

116
00:11:47,360 --> 00:11:49,760
Well, there's a good reason for that.

117
00:11:50,060 --> 00:11:55,790
That is that next time when I subdivide the thing in the middle is going to be ambiguous depending on which way you look.

118
00:11:56,270 --> 00:12:02,630
Whereas if you take this rule is consistent all the way through, okay, well, that's what you do.

119
00:12:03,530 --> 00:12:07,820
And if you do that, you get a pattern like this.

120
00:12:08,150 --> 00:12:20,580
Well, let's do it. Okay. I'm going to start with a nice big painting in there, and then I'm going to subdivide that according to the rules.

121
00:12:20,580 --> 00:12:24,920
And I've just been giving you the volume up three.

122
00:12:26,810 --> 00:12:34,490
Is that? Do I just shout or is there a way of doing it with. Oh, it's down there.

123
00:12:36,920 --> 00:12:42,080
Oh, another one. It's not going to interfere with the other one. Awful squawks and things like that.

124
00:12:42,770 --> 00:12:45,950
Good. If it squawks out, I'll hurriedly take one out of pocket.

125
00:12:46,760 --> 00:12:50,690
Can you hear me now? Good. Okay.

126
00:12:51,020 --> 00:12:55,190
So that's the Pentagon before I subdivide that.

127
00:12:55,850 --> 00:12:59,330
There we go. Just according to the rules I was giving you.

128
00:12:59,960 --> 00:13:04,490
And then I subdivide that. No.

129
00:13:06,930 --> 00:13:12,060
Now, you see, there would have been this little ambiguity possibly here because there's that rhombus.

130
00:13:13,800 --> 00:13:19,030
And I look to see where should I put the Pentagon?

131
00:13:19,050 --> 00:13:23,490
Well, I look around and I say, there is that pattern of Pentagons which I want you you would be able to find.

132
00:13:23,820 --> 00:13:26,850
And I look at that, that one's over there has to be there.

133
00:13:27,150 --> 00:13:31,320
And therefore this one's down there. Okay. So that's satisfying the rules.

134
00:13:32,850 --> 00:13:39,960
And then next time, once more and I get the pattern that we had before.

135
00:13:40,410 --> 00:13:43,530
Okay. So that tells you how it's built.

136
00:13:43,980 --> 00:13:47,670
It's got hierarchical construction that can whip those away.

137
00:13:48,040 --> 00:13:52,650
And just to confuse you a bit, I'm turning it upside down. Now, where was that original big Pentagon?

138
00:13:54,450 --> 00:13:58,560
It's not so easy to find, but I. Each time I had to figure it out all over again.

139
00:13:59,250 --> 00:14:06,180
The point about this is that it has a uniformity that's not so obviously hierarchical.

140
00:14:06,330 --> 00:14:12,270
It does have a hierarchical construction, but it's really something much more regular than that.

141
00:14:12,360 --> 00:14:18,090
It has a highly irregular structure, and the hierarchical nature is not at all obvious.

142
00:14:19,710 --> 00:14:23,610
In fact, it has various properties. I point out a few of these.

143
00:14:24,420 --> 00:14:28,410
For example, you see here you have a regular deck again, ten sided figure.

144
00:14:29,040 --> 00:14:34,830
There's another one. And every time when you find one of these regular deck of guns, it's always subdivided in the same way.

145
00:14:35,160 --> 00:14:39,000
Three pentagons to rhombus and one pentacle.

146
00:14:39,030 --> 00:14:42,660
That one just as cap. Okay.

147
00:14:42,840 --> 00:14:48,570
And every time you find one of these, it's always surrounded by a ring of ten pentagons.

148
00:14:48,960 --> 00:14:53,520
Yeah, there's another one. Always a ring of ten pentagons, wherever it is.

149
00:14:53,700 --> 00:14:57,329
Sometimes these things overlap. Like here you've got two of them overlapping.

150
00:14:57,330 --> 00:15:02,640
Another one there. Whenever that happens, you still find your ring of ten pentagons that just go quite happily through each other.

151
00:15:03,390 --> 00:15:06,600
And it has that very nice property.

152
00:15:07,260 --> 00:15:16,049
Other things which are very evident to me just while I stand here, is that if I take any line in the picture, I can put a ruler against it.

153
00:15:16,050 --> 00:15:26,820
So any line in the picture, it will keep on going right across the picture with the same density of lines all the way up.

154
00:15:27,690 --> 00:15:34,590
So if you imagine a boat, you had a field of corn and there was one growing at each point vertex of this pattern.

155
00:15:34,920 --> 00:15:39,670
And you drive past in the car and you look at it. They all line up at some point and then it all line up again.

156
00:15:39,690 --> 00:15:43,019
You say, Oh, that's a nice regular pattern. Well, it's one of these things.

157
00:15:43,020 --> 00:15:48,150
Maybe this you'd be lucky. Lucky to find a farmer who'd do that, but never mind.

158
00:15:49,200 --> 00:15:52,799
Anyway, I just wanted to indicate that there is a lot of structure.

159
00:15:52,800 --> 00:15:59,160
There's a lot of a lot of there are a lot of features which are not at all evident from the hierarchical construction.

160
00:15:59,790 --> 00:16:05,309
And for example, now we have this ten sided one in the middle here with its ring of Pentagons.

161
00:16:05,310 --> 00:16:11,459
But in this case, you also find right the way around it a nice ring where you have rhombus, Pentagon,

162
00:16:11,460 --> 00:16:16,170
rhombus, Pentagon, rhombus, spending all the way around ten, ten, rhombus and pentagons alternating.

163
00:16:16,920 --> 00:16:20,220
And you find those all over the place. Okay.

164
00:16:21,900 --> 00:16:33,150
Now, I want to say something else about these patterns, and that is there was a problem originally to do with tiling, playing with squares,

165
00:16:33,540 --> 00:16:40,530
with coloured edges, and you had different squares with different arrangements of colours and you had to match the colours.

166
00:16:41,310 --> 00:16:49,980
And the question that was raised by a Chinese-American mathematician called How one could you find a

167
00:16:49,980 --> 00:16:56,850
computer program which will answer yes or no given those sets of colourings finite set of the squares?

168
00:16:57,360 --> 00:17:02,040
Yes or no? Can you tile the entire plain with these coloured squares?

169
00:17:02,970 --> 00:17:09,300
And he came to the conclusion that yes, there would be an algorithm for doing that, a computer program for doing that,

170
00:17:09,900 --> 00:17:19,620
if it were the case that for any set of coloured squares, if it did tile a plain, you could be sure that it would tile a plane in a periodic way.

171
00:17:21,150 --> 00:17:23,220
Well, his student, Robert Berger,

172
00:17:23,700 --> 00:17:33,360
thought about this problem and eventually proved that there was no computer program for solving the tiling problem of this kind.

173
00:17:33,900 --> 00:17:48,930
And therefore, as part of his argument, in fact, he produced a set of, I think it was originally 2000, 20,426 different coloured squares.

174
00:17:50,760 --> 00:17:56,310
He then got it down to about 100 and then Rafi Robinson got it down to about six.

175
00:17:57,210 --> 00:18:02,180
And he had a way of making it into a jigsaw puzzle rather badly drawn here, but that's what it is.

176
00:18:02,190 --> 00:18:07,110
You can see the squares on this piece of paper, but that's not got the designs on them.

177
00:18:07,800 --> 00:18:12,720
You have to fit these together and the only way you can do it is known periodically.

178
00:18:13,290 --> 00:18:18,630
So that was made it consistent that there was no computer program for solving the tiling problem

179
00:18:18,900 --> 00:18:27,060
with these coloured squares and the fact that there are shapes which are only known periodically.

180
00:18:27,840 --> 00:18:36,660
Well, I was visited in the Old Mathematical Institute by the American mathematician Simon Kitchen, and he was describing these things to me.

181
00:18:37,080 --> 00:18:41,310
And he said and Robinson was somebody like to get the number down as small as possible.

182
00:18:42,150 --> 00:18:46,860
And so he told me there were six. And I said, Well, I think I can do it with five.

183
00:18:49,140 --> 00:18:54,030
And my five will look completely different from his. In fact, what I was actually doing was with six.

184
00:18:54,360 --> 00:19:00,180
Now, those six, you see, are actually the same basically as this pattern.

185
00:19:01,590 --> 00:19:07,499
However, the Pentagon's are divided into three classes, depending upon well,

186
00:19:07,500 --> 00:19:12,330
as with this one, we have the Pentagon surrounded by five of us here we have it only three.

187
00:19:12,480 --> 00:19:15,640
And here we have another one. Only two here.

188
00:19:15,660 --> 00:19:20,670
So either two, three or five, and I'm going to make those different.

189
00:19:21,570 --> 00:19:26,700
And then I make it into a jigsaw puzzle where these are the that's the five.

190
00:19:26,700 --> 00:19:33,600
That's the two. The three and that's the two one. And just to show you what's going on, a couple of these so that.

191
00:19:34,760 --> 00:19:41,210
The five one is the is the orange one. And you can see what's goes on.

192
00:19:43,820 --> 00:19:50,700
Here we go. But you see, that's still six like Rafael Robinson six.

193
00:19:51,790 --> 00:19:57,989
But but yeah. So if you imagine the coloured ones are different and they have a matching rule according to those

194
00:19:57,990 --> 00:20:04,410
little knobs and matches which I just showed you then the point is that you force that kind of pattern,

195
00:20:04,410 --> 00:20:08,760
which I've been describing. It does have this hierarchical arrangement, but I'm not saying do it that way.

196
00:20:08,970 --> 00:20:13,470
I'm saying do it anyway you can where you make the notches and the knobs fit together.

197
00:20:14,400 --> 00:20:18,270
And the only way you can do it is in the sort of arrangement which I just showed you.

198
00:20:19,050 --> 00:20:24,540
Now, the reason I said I thought I could do it with four.

199
00:20:24,930 --> 00:20:32,460
You see, if you look at these tiles, you see that there's this little funny star shape thing there and two of them on here.

200
00:20:32,580 --> 00:20:35,700
And there's only one little place where you have a a slot for them to go into.

201
00:20:35,700 --> 00:20:38,880
And that's one this one. So you cut this one out and glue it to that.

202
00:20:39,150 --> 00:20:43,200
Take another two versions of this, glue it on here and here, and you don't need that one.

203
00:20:43,650 --> 00:20:49,380
So you could do it with five. So I went home and then I started thinking about this, cutting up and seeing whether I should improve on that.

204
00:20:49,680 --> 00:20:57,390
And I got it down to four and I was quite pleased with that. And then I got it down to two and I wasn't I wasn't so pleased.

205
00:20:58,080 --> 00:21:03,270
Now, why wasn't I so pleased? I think the reason that I wasn't so pleased was I thought, Oh, this is so easy.

206
00:21:03,960 --> 00:21:08,160
It must have been known to people. I think that's why I thought that.

207
00:21:08,160 --> 00:21:12,630
But I do remember being disappointed when it came down to I'd rather like my set of four,

208
00:21:13,110 --> 00:21:20,400
but instead of two I originally hit on is shown in this picture here where we have.

209
00:21:22,810 --> 00:21:29,420
I well, it's cuts and dots. Here we have the cut and here we have the dart and those are the only two.

210
00:21:29,750 --> 00:21:36,290
And I have a rule where they have to they've got corners which are either coloured black or white, and you have to match the corners.

211
00:21:36,410 --> 00:21:40,180
You can do this with knobs and notches, if you like, make a jigsaw puzzle out of it.

212
00:21:40,180 --> 00:21:49,460
And well, I'll show you one in a minute. And each one of the cuts you have to put lines on in the appropriate way, and each dart you do as well.

213
00:21:50,090 --> 00:21:58,580
And I've done that on this side, on the right hand side, and you will see that it brings out the pattern we just had before with the Pentagon's.

214
00:21:59,910 --> 00:22:05,010
It's just they assembled to make the Pentagon's and the the dotted ones the shaded

215
00:22:05,010 --> 00:22:10,050
ones make either the rhombus here or the just as cap here or the pentacle over there.

216
00:22:10,470 --> 00:22:19,070
And that is the pattern we just had before. And you can see that if I do it right, then watch.

217
00:22:19,080 --> 00:22:22,080
Or which one I need that. Let's try that.

218
00:22:23,100 --> 00:22:26,430
And probably East Orange one.

219
00:22:27,660 --> 00:22:31,050
Let's try that one. Yeah, I think that's it.

220
00:22:33,530 --> 00:22:42,920
So you can see they're really the same kind of pattern. But the cuts and cuts are done in a way that you don't need so many different styles.

221
00:22:42,980 --> 00:22:49,910
That's all. Now, you can also modify them, as I say, to make a jigsaw puzzle type arrangements.

222
00:22:50,360 --> 00:22:59,840
And here's one which is sort of influenced by Escher to some degree, where there are two birds, the big bird and the little bird.

223
00:23:00,290 --> 00:23:06,830
And the only way you can assemble them is a jigsaw puzzle, is in an arrangement like this which never repeats itself.

224
00:23:07,640 --> 00:23:13,390
And you can then go back and see how that ties in with the these ones.

225
00:23:13,400 --> 00:23:21,080
And I hope I can. It wasn't designed to fit this transparency, but I managed to find a place where they fit.

226
00:23:21,080 --> 00:23:27,650
And there I think you can see the big birds, the cuts and the little birds are the dots.

227
00:23:29,670 --> 00:23:34,650
And that's all they are. But that makes a nice jigsaw puzzle.

228
00:23:36,060 --> 00:23:44,730
Okay. Now, you can also do two in another way, which are actually more commonly used.

229
00:23:44,770 --> 00:23:46,290
It turns out because they're just rhombus.

230
00:23:47,010 --> 00:23:55,680
And the colouring I've done here is you have to match those colours and it forces this non periodic arrangement which I've just been showing you.

231
00:23:56,160 --> 00:24:02,080
And what you get, I'll start off by relating it to the cats and dogs.

232
00:24:03,930 --> 00:24:16,310
There you are. So each. Each dot is drawing, you have the same lines drawn on in each cut the same way and up with this pattern of promises.

233
00:24:18,440 --> 00:24:25,969
Now, this pattern has been used in all sorts of places. I'll show you some places where it has been used, but often just like this.

234
00:24:25,970 --> 00:24:29,750
And I always thought that's a bit disappointing because you don't know why the pattern is like that.

235
00:24:29,930 --> 00:24:33,200
You think, well, lots of ways of doing wrong business. You could just make them regular patterns.

236
00:24:33,200 --> 00:24:39,230
So why do you do them like this? And it's you might say, well, that's what you're supposed to do, but that's not good enough, is it?

237
00:24:40,130 --> 00:24:46,220
But if you have the matching rules like here, then it does force that type of arrangement.

238
00:24:47,450 --> 00:24:57,140
I use a few things I should say about this. One is okay as it forces it, but are there more than one way of doing it?

239
00:24:57,740 --> 00:25:02,360
Is they want more than one way of doing it? Well, it depends.

240
00:25:02,600 --> 00:25:09,710
The answer is both. It's unique and there's infinitely many different ways of doing it.

241
00:25:10,010 --> 00:25:16,220
In fact, true to the left, note the number of points that are on the on the real line two to the left note number of different ways of doing it,

242
00:25:17,090 --> 00:25:22,400
but they're all the same in a certain sense. If you were given two of these patterns and you wanted to see whether they were the same or different,

243
00:25:23,000 --> 00:25:27,440
and you could only examine a finite portion you would never tell because any

244
00:25:27,440 --> 00:25:32,240
finite portion in one of them will appear in the other one infinitely many times.

245
00:25:32,930 --> 00:25:37,790
So as far as finite arrangements are concerned, they're identical.

246
00:25:38,330 --> 00:25:43,010
The only difference is our way to infinity. And that's quite a subtle thing.

247
00:25:44,660 --> 00:25:50,750
Well, let me not go into the subtleties of that, but it is the case that, strictly speaking,

248
00:25:50,750 --> 00:25:54,320
as far if you go all out to infinity, there, all the different ways of doing it.

249
00:25:54,830 --> 00:26:01,160
It depends on where you start. In the hierarchies, there are hierarchical, hierarchical in the same way that I started,

250
00:26:01,550 --> 00:26:06,320
but you could sort of start in different places, and that makes a difference, such as infinity.

251
00:26:06,320 --> 00:26:09,500
But as far as financial arrangements are concerned, they're all the same.

252
00:26:10,280 --> 00:26:17,309
Okay. Now. There are these are all the five fold ones or ten fold.

253
00:26:17,310 --> 00:26:21,330
It depends whether you you can call them either five fold or ten fold, depending on your rules.

254
00:26:21,720 --> 00:26:25,830
I'm not going to make a distinction there, but there are others.

255
00:26:26,070 --> 00:26:32,130
And here we have a 12 fold, one that's rather nice. One made by Galen Dennison.

256
00:26:33,990 --> 00:26:36,330
He's Swiss. I'm not sure it was Swiss.

257
00:26:38,850 --> 00:26:43,110
These were there was a competition, actually, for the I thought the ones that they produce were the nicest ones.

258
00:26:43,230 --> 00:26:49,350
These are 12. So that's one of them. And they haven't got matching rows, but you can make matching rules.

259
00:26:49,500 --> 00:26:54,900
It's a little bit more complicated how to do it. You can't do it with just I think there are three different shapes here.

260
00:26:55,590 --> 00:26:59,579
But you can you have to have another shape to make to force the matching rules.

261
00:26:59,580 --> 00:27:04,800
It's really a four tile thing. Here's another version of the same thing.

262
00:27:04,920 --> 00:27:11,640
Really rather nice. There are also the eight fold ones which Robert Mann and somebody could beat to produce.

263
00:27:14,010 --> 00:27:15,750
Very rapid momentum is very rapid.

264
00:27:16,230 --> 00:27:24,720
Martin Gardner had an article on these things and he announced he was going to produce these something that he hadn't ever shown before.

265
00:27:25,230 --> 00:27:28,260
And I don't think he said very much about what they were like.

266
00:27:28,680 --> 00:27:40,470
And Robert Manne, this rediscovered the The Rumpus ones, which is very remarkable I thought in less than a month, I suppose.

267
00:27:41,850 --> 00:27:45,480
Okay, now what do I want to say?

268
00:27:49,750 --> 00:27:55,030
Well, sometimes people say, well, you find these things in Islamic patterns, Islamic buildings,

269
00:27:55,480 --> 00:28:01,180
and you certainly find a lot of extraordinary looking patterns in Islamic buildings with all sorts of symmetries.

270
00:28:01,750 --> 00:28:05,050
Yes, it's nice example here. And you see this part of the wall.

271
00:28:06,970 --> 00:28:10,920
I'm not sure I've counted what these ones. So that's ten fold, I think, here and so on.

272
00:28:10,930 --> 00:28:16,540
But they're all limited regions and there's no rule about how you might go out to infinity.

273
00:28:16,540 --> 00:28:24,250
Or if you see a nice pattern of these things, it's almost always the case that there's a evident symmetry to it, that it just repeats itself.

274
00:28:24,910 --> 00:28:30,069
And so just having interesting symmetries, I mean, the symmetries don't extend to the whole pattern.

275
00:28:30,070 --> 00:28:31,360
They just local symmetries.

276
00:28:31,930 --> 00:28:38,590
So you can have these local symmetries which are not crystallographic, but the pattern as a whole is a crystallographic thing.

277
00:28:39,550 --> 00:28:44,980
One of the more remarkable examples was pointed out by a beach map.

278
00:28:45,010 --> 00:28:49,090
And this. No, I said that's wrong.

279
00:28:49,270 --> 00:28:53,720
That's it. Yes. And Paul Steinhart and a Chinese chap.

280
00:28:53,740 --> 00:28:59,350
I've gotten his name now. But anyway, they did try and relate it to the tilings that I had.

281
00:28:59,980 --> 00:29:07,570
I think there are some relationships, but. But there's no evidence that there was any kind of a rule enforcement periodicity.

282
00:29:07,570 --> 00:29:13,000
In fact, the examples like the one up there you can see actually is is completely periodic.

283
00:29:14,580 --> 00:29:18,050
So and it's less obvious here what's going on.

284
00:29:18,060 --> 00:29:24,750
And they have a certain hierarchical aspect. You see, they're subdivided, but not in a way that repeats the patterns on the big scale.

285
00:29:24,840 --> 00:29:30,000
So it's not clear whether that may be somewhere in some building.

286
00:29:30,430 --> 00:29:34,739
A such an example will actually be found of these non periodic things.

287
00:29:34,740 --> 00:29:43,830
But I want to show you something else which is a bit different somewhat later, but not that much later.

288
00:29:44,460 --> 00:29:58,680
1619. These are pictures done under the instruction of Johannes Kepler, the famous astronomer who discovered the Kepler in orbits, ellipses and so on.

289
00:29:59,460 --> 00:30:08,760
But he also fooled around with all sorts of things and he like polyhedron non none and how you fill space with different shapes and so on like that.

290
00:30:09,360 --> 00:30:21,690
But he had a page in one of his, his books, harmonica mundi volume two 1619 in which there are all these different, very curious symmetries.

291
00:30:23,040 --> 00:30:34,170
Now, my father owned a copy of this book, and I had seen this page some years before finding these patterns, which I've just been showing you.

292
00:30:34,980 --> 00:30:39,150
But I wasn't thinking about them at the time. But let me show you something.

293
00:30:40,860 --> 00:30:45,120
I'm going to concentrate now on this person called a one there.

294
00:30:45,780 --> 00:30:51,720
Let's make it a bit bigger. There we are. And now I want to show you something.

295
00:30:52,300 --> 00:30:57,870
This is made of pentagons, regular pentagons put together. And then these little holes in them which are decaying in all shapes.

296
00:30:57,870 --> 00:31:04,380
And these stars, these pentacles. Now, I'm going to try and find the right spot here, which is maybe a little tricky.

297
00:31:04,470 --> 00:31:08,040
Here we are. And look, it fits.

298
00:31:08,040 --> 00:31:15,600
Exactly. Not only does it fit exactly, but in Kepler's picture, he drew an extra little line there for some reason.

299
00:31:16,320 --> 00:31:19,500
And that line also fits. Exactly.

300
00:31:20,670 --> 00:31:25,590
Now. What was he doing? I don't know. I really don't know.

301
00:31:26,820 --> 00:31:30,660
I was giving a lecture like this somewhere, and somebody in the audience at the end said, yes.

302
00:31:30,660 --> 00:31:36,360
There is a letter that Kepler wrote to somebody and he explained what he was doing in his pictures.

303
00:31:38,250 --> 00:31:42,720
I said, Oh, that's very interesting. Can you find out and let me know? Well, I heard nothing more from her.

304
00:31:43,980 --> 00:31:47,070
There's times I've got people to look at Kepler's letters and things,

305
00:31:47,070 --> 00:31:52,649
and the best they could find were letters that he had written to the person who actually did the drawing.

306
00:31:52,650 --> 00:31:59,040
So that was a he had a an artist actually to draw these things for him, but were strictly under his instruction.

307
00:31:59,520 --> 00:32:01,050
So they were Kepler's pictures, no doubt.

308
00:32:02,160 --> 00:32:10,680
My guess is that he was probably interested in he certainly interested in face space filling with with regular shapes and so on and crystals.

309
00:32:10,680 --> 00:32:12,989
But people didn't even know really about atoms in those days.

310
00:32:12,990 --> 00:32:23,490
But it was one possibility and I imagine that the idea of crystals being atoms arranged in a regular way was certainly a strong possibility.

311
00:32:24,060 --> 00:32:34,410
And I suspect that Kepler was also interested in living things and maybe plants and so on might have had made use of some other symmetries,

312
00:32:34,410 --> 00:32:38,190
such as the ones in his pictures here. I have no idea.

313
00:32:38,190 --> 00:32:47,910
That's pure speculation. There was a time when people were just finding examples of what we call quasicrystals,

314
00:32:48,090 --> 00:32:54,930
which seem to have a ten fold, five fold and ten fold symmetries, and then they seem to find some eight fold ones.

315
00:32:55,590 --> 00:33:06,900
And then I was visiting Switzerland. This chap, Nissen, did one of the pictures I showed you earlier, the 12th of one, he and a colleague.

316
00:33:07,350 --> 00:33:15,120
This these ones. And the reason he was interested in this was that he claimed to have an alloy which

317
00:33:15,120 --> 00:33:19,840
produced diffraction patterns which seem to suggest there were 12 fold symmetry.

318
00:33:20,700 --> 00:33:25,350
And nobody seemed to believe him when he showed me these things. And I thought, well, it looks fairly persuasive.

319
00:33:25,700 --> 00:33:31,109
They weren't quite uniform over the whole picture. And then he showed me the data points from the diffraction pattern.

320
00:33:31,110 --> 00:33:35,220
These are electrons you shine electrons on and they scatter out at particular angles.

321
00:33:35,490 --> 00:33:41,700
And then you see the shapes of the the little points of light that you get where the electrons hit.

322
00:33:42,270 --> 00:33:48,590
And this tells you something of the symmetry. And he showed me this with a 12 fold symmetry, little points.

323
00:33:49,060 --> 00:33:53,110
I looked at this and I started joining them up with the squares of triangles and like that.

324
00:33:53,160 --> 00:33:56,190
Now I've seen that picture before somewhere.

325
00:33:57,330 --> 00:34:01,830
And where on earth have I seen that before? And then it dawned on me.

326
00:34:02,340 --> 00:34:04,670
Yeah. It's that one.

327
00:34:05,390 --> 00:34:15,960
If you put the points of the diffraction pattern on the corners of this arrangement here, if it is a bit bigger, that's the diffraction pattern.

328
00:34:15,980 --> 00:34:20,360
Well, I published some other points scattered about, but it was it was that particular pattern.

329
00:34:20,930 --> 00:34:22,850
Again, I have no idea what Kepler was doing.

330
00:34:23,420 --> 00:34:30,860
I'm sure he didn't know anything about diffraction patterns and those, but nevertheless, he had some deep insights in other ways, too.

331
00:34:30,890 --> 00:34:34,850
So who knows? I'll just leave you. That's a bit of a conundrum.

332
00:34:36,110 --> 00:34:43,040
Okay. Let's do something a little different here. If I can get it to work, which is not necessarily the case here.

333
00:34:43,040 --> 00:34:46,459
We have I'm afraid it's got it's an old slide and you'll see the got rather messed up.

334
00:34:46,460 --> 00:34:50,540
I must get a new one. But these are composites, which you can probably just about see.

335
00:34:50,540 --> 00:34:55,940
But there there's a nice big pattern of the rhombus. I can do it with that, can't I?

336
00:34:56,510 --> 00:35:01,010
Yeah. I don't need the transparencies. Okay.

337
00:35:01,040 --> 00:35:04,840
So that promises. What's that like?

338
00:35:06,610 --> 00:35:10,480
Well, that's wrong. As is, too. But I think I can't even see them very well.

339
00:35:10,780 --> 00:35:19,660
I can see those blobs there. But that little rhombus, this point about this is you can put others on top and get nice diffraction patterns.

340
00:35:21,790 --> 00:35:26,590
I have to worry about that spot, because that's just the light, isn't it? That's.

341
00:35:27,830 --> 00:35:33,590
Probably do. Never mind. Don't worry about that. Um.

342
00:35:34,910 --> 00:35:40,330
I don't know whether I can do this very well here. But you can find a little place where.

343
00:35:41,080 --> 00:35:52,240
Oh. That's good. See if I can get that spot in the middle now.

344
00:35:53,800 --> 00:36:14,220
And a nice one. If you see those lines, the lines are where they disagree and where you don't see a line.

345
00:36:14,220 --> 00:36:18,090
That's where they agree. Now, this is where she'll probably fail.

346
00:36:18,090 --> 00:36:23,820
But let me try it. I've marked a spot where.

347
00:36:24,930 --> 00:36:31,110
Let's just see. That's not it. I thought I knocked it.

348
00:36:34,080 --> 00:36:53,540
To see if. Okay. If you get it right, which I.

349
00:36:53,550 --> 00:37:01,630
Which is hard to do, I'm afraid. I think it's best if a twisted a bit more.

350
00:37:18,870 --> 00:37:26,360
That's confusing that on the one. It's not working.

351
00:37:43,820 --> 00:37:50,060
Let's try and match my marks. That might help. No.

352
00:37:50,900 --> 00:37:51,920
It's not working very well.

353
00:37:52,760 --> 00:38:00,200
There should be just oh, yes, there's a line right down the middle and that's it matches everywhere except along that line.

354
00:38:02,330 --> 00:38:08,050
Uh. Not very good.

355
00:38:21,340 --> 00:38:26,020
I think it's best that you do. You can probably see the line extending from here to here.

356
00:38:26,350 --> 00:38:32,040
That line is where it disagrees. Three degrees everywhere else.

357
00:38:32,040 --> 00:38:36,420
You see, that's one of these 99.9 things, if I got it right.

358
00:38:43,220 --> 00:38:57,290
I think I shouldn't wait too much time on that. It ought to be just about there.

359
00:39:07,990 --> 00:39:13,260
I think you just about it. Actually glue them together and then it.

360
00:39:13,260 --> 00:39:21,030
But of course, the defect isn't. I think that's the best I'm going to be able to do.

361
00:39:21,040 --> 00:39:25,630
I'm sorry about it. I think one trouble also is that these machines.

362
00:39:27,120 --> 00:39:32,580
Well, I shouldn't blame it on the machine. It's not quite.

363
00:39:35,420 --> 00:39:44,520
Did you get it today or. So I took so long.

364
00:39:46,200 --> 00:39:50,040
I think perhaps I'll use this other machine. Oh, I have to get it going, don't I?

365
00:39:50,730 --> 00:39:53,730
I know how to move the picture, but do I turn it? How do I turn it on?

366
00:39:54,690 --> 00:40:06,190
Is it there? What is that square? No, it's not it. I'm very stupid at these things.

367
00:40:08,240 --> 00:40:12,900
Should I be pressing something back front over there?

368
00:40:12,930 --> 00:40:17,420
Of course. Perhaps it was there all the time, was it? Yes.

369
00:40:18,050 --> 00:40:21,620
Well, you can see this actually hasn't been put up in the building yet.

370
00:40:21,860 --> 00:40:27,410
This is a nice poster design in our old building. I think it's being stored somewhere away, isn't it?

371
00:40:27,950 --> 00:40:36,530
Probably. Anyway, it's made up out of little pieces, plastic pieces that were given to me by a mathematician, Ron Graham.

372
00:40:37,010 --> 00:40:40,550
And some of them were built by. Made by some other people.

373
00:40:41,700 --> 00:40:44,980
Michael. What? His company and.

374
00:40:46,580 --> 00:40:54,470
The big one. The middle is the wrong one, and they're basically cats and dogs with the matching rules done by little knobs and things.

375
00:40:54,920 --> 00:41:02,630
And it's also coloured in a very specific way, which the colouring is unique up to permutation of the colours,

376
00:41:04,130 --> 00:41:09,260
and it produces some nice things if you sort of squint your eye and look it up on the top left,

377
00:41:09,260 --> 00:41:13,219
we see a nice, straightforward, got dots on the top right, rhombus,

378
00:41:13,220 --> 00:41:21,770
one bottom left we see the modification which gives you the birds and the relationship to the Pentagon's as well.

379
00:41:22,190 --> 00:41:28,100
So that's a little bit baffling. Right in the middle there is a a dog and the dog is not.

380
00:41:28,700 --> 00:41:35,839
It's cheating. He's a different shape from the others. But if you have one dog in there, the tiling is completely unique, right out to infinity.

381
00:41:35,840 --> 00:41:40,070
So you don't have this business of up to the left, not that unique.

382
00:41:41,390 --> 00:41:46,470
And the one thing on the right with the diffraction pattern is an actual quasicrystal.

383
00:41:46,490 --> 00:41:53,090
This is an actual material produced by people in Japan and now you can produce ones that you can actually see.

384
00:41:53,090 --> 00:42:00,559
I mean, it's probably only about that big, but you can see it quite clearly and it's a regular dodecahedron which is not allowed for a crystal.

385
00:42:00,560 --> 00:42:09,410
So it's quite interesting that you can have that. And there's a bigger picture of the actual quasi crystal.

386
00:42:09,890 --> 00:42:17,000
I think they're called crystals now. So I don't know. I think they have so many properties of crystals that people prefer to call them all crystals.

387
00:42:18,040 --> 00:42:27,519
I'm not sure whether I like that or not. Never mind. No part of the fact that's upside down, which maybe shouldn't matter.

388
00:42:27,520 --> 00:42:37,870
It's a hit. I hope not at all upside down. That is the first time these things were ever used in architectural design.

389
00:42:37,990 --> 00:42:44,799
That was anywhere. Well, I hope the others are not like that. An architectural design on a building.

390
00:42:44,800 --> 00:42:52,750
This was the Tokyo Metropolitan University and the architect there wanted to use clocks dart so you

391
00:42:52,750 --> 00:42:57,100
could probably see them a little bit hidden with having all sort of other marks on top of them,

392
00:42:57,670 --> 00:43:05,590
but cutting dots. And when the thing was assembled, the architect looked at it very carefully and he found that it made a mistake.

393
00:43:06,010 --> 00:43:09,970
And had they they hadn't take it out and put it back correctly, which impressed me a lot.

394
00:43:10,480 --> 00:43:18,340
It's completely right. This, as far as I can tell, it's supposed to have some connection with the shape of Tokyo as well,

395
00:43:18,340 --> 00:43:22,400
which I don't know much about, but I think it was rather very nice.

396
00:43:22,570 --> 00:43:30,570
Nicely done. Okay. That's right. Way up. This is a rather gaudy looking building in Melbourne.

397
00:43:30,570 --> 00:43:40,350
I think it is Australia, in Melbourne where they've used quite enough all over the front and all over and inside too.

398
00:43:40,350 --> 00:43:46,710
And it's it's rather an extraordinary place. It's one of the first buildings I know which made use of these things.

399
00:43:47,550 --> 00:43:55,410
And you can see up on the top, right. I think that's got the stripes on them and I was showing you, but it seemed to be pretty accurately done.

400
00:43:55,740 --> 00:43:59,250
They've got the stripes on the other ones too. Yeah. Which you have to match.

401
00:44:00,830 --> 00:44:05,300
Now there's a much more sedate version on the ground down here.

402
00:44:05,810 --> 00:44:14,840
You see, this is just straightforward robberies. This is Stony Brook in Long Island, New York, and that's the mathematics building.

403
00:44:14,840 --> 00:44:19,460
And they wanted to use a tiling of this kind. But it's just straightforward promises.

404
00:44:20,540 --> 00:44:24,170
As you'll find a lot of see this against dreadful promises.

405
00:44:24,410 --> 00:44:33,260
This is in in Perth, in Australia and I think this is the Cosmology Centre or something.

406
00:44:33,260 --> 00:44:43,280
It's a very nice popular science place in, in Perth and they wanted to use the tiling from the starlings on the floor.

407
00:44:44,000 --> 00:44:49,610
Perfectly nicely done. I think the biggest area I've ever seen is also in Perth.

408
00:44:49,820 --> 00:44:53,750
It is in their chemistry building and again it's just straight run business.

409
00:44:55,940 --> 00:45:02,130
And here we have cut and. This is. I think. In the United States.

410
00:45:05,120 --> 00:45:07,520
Carleton. Carleton University somewhere. Yep.

411
00:45:08,120 --> 00:45:17,780
Which is a nice done touch and ask without also bringing out some pattern in the colouring which is very attractive.

412
00:45:20,070 --> 00:45:27,270
This is the other place. St John's College in Cambridge.

413
00:45:27,990 --> 00:45:31,440
And this is the entrance to the building was called the Penrose building.

414
00:45:31,450 --> 00:45:39,870
And I say it's nothing to do with me whatsoever that some architect who designed buildings at St John's College in the 19th

415
00:45:39,870 --> 00:45:47,490
century and it was converted to become the library building and for the entrance they wanted to put this door swings round.

416
00:45:47,500 --> 00:45:47,910
You see,

417
00:45:48,240 --> 00:45:55,290
the door is now open so you can go either side when it's closed that you can only see half the pattern and the door swings round and it's cuts.

418
00:45:55,290 --> 00:46:02,510
And that's perfectly correctly done. And here we have what we call a judgement.

419
00:46:03,680 --> 00:46:09,470
This is a pattern which for it's just in front of the the.

420
00:46:11,010 --> 00:46:18,870
In front of the student bar and it's usually full filled with lots of cigarette stubs and beer stains and so on.

421
00:46:18,980 --> 00:46:27,030
But I think it's been reasonably scrubbed down, but not terribly perfectly, and it is matching rules.

422
00:46:28,200 --> 00:46:33,479
So I was with my wife where they were laying this thing, and I went to a play at the Playhouse,

423
00:46:33,480 --> 00:46:37,650
I think, and we came back and they were just about finishing the laying down.

424
00:46:37,650 --> 00:46:43,320
And I thought, well then have a look and see what they've done. And so I looked at it and all right, I thought, but a little bit disturbing.

425
00:46:43,910 --> 00:46:48,750
And so I went up on the higher level to look down on it. And I kept thinking, there's something a bit disturbing about that.

426
00:46:49,350 --> 00:46:55,230
And then I realised right at the edge one of the workmen had seen you could put another tile in which would fit alright,

427
00:46:55,830 --> 00:47:02,490
but it would fit alright there. But if you kept on going somewhere in the middle of the lawn, you'd find you get stuck.

428
00:47:04,050 --> 00:47:07,590
So I had to have it pull it out, I'm afraid. And what's there now is correct.

429
00:47:07,590 --> 00:47:12,770
But. But there is this danger because it is a non-local business.

430
00:47:14,640 --> 00:47:20,440
Let me move on to. Yes.

431
00:47:20,440 --> 00:47:25,030
This is actually a three dimensional version. This is, I hope, going to appear on the wall.

432
00:47:25,780 --> 00:47:35,200
The tiling out front of our building was financed by my Andre Stern very generously,

433
00:47:35,620 --> 00:47:39,730
and he wanted having a new building made, and he wanted to have something in there a bit different.

434
00:47:40,300 --> 00:47:47,680
So I suggested this three dimensional version. It's just the rhombus, but you see that they're all the same size and shape.

435
00:47:47,680 --> 00:47:52,090
The rhombus is just the angle that you see that makes the either the fat one or the thin one.

436
00:47:52,690 --> 00:47:56,440
And it makes this undulating terrain, which is quite nice.

437
00:47:59,980 --> 00:48:08,050
And here we have our building. So let me try and say something about this now, because it is the rhombus tiling.

438
00:48:09,340 --> 00:48:13,270
But it's got more to it. And I want to describe.

439
00:48:13,390 --> 00:48:16,700
I'm afraid I shot this up on that. There is the rhombus tiling.

440
00:48:16,720 --> 00:48:27,160
But you see more to it there. There's some patterns there. Now, you may remember that when I was showing the pattern of pentagons.

441
00:48:27,270 --> 00:48:30,520
So, um, let's take the bigger version here.

442
00:48:31,510 --> 00:48:35,560
You have places where there are these regular decorations.

443
00:48:36,160 --> 00:48:41,260
And each time you have regular again, it's surrounded by a ring of ten pentagons.

444
00:48:42,190 --> 00:48:47,740
So I thought maybe you could enhance these rings somewhat by having a circle going around them.

445
00:48:49,780 --> 00:49:00,130
So there you are. Now, that's, of course, doesn't do the same thing to all the Pentagon's.

446
00:49:00,910 --> 00:49:06,340
So let's do something else. I'm going to complete these things a little bit.

447
00:49:09,000 --> 00:49:17,170
If I do that, you can see that it's really a sort of fattened up version of the the.

448
00:49:17,680 --> 00:49:22,450
Here's the Pentagon fattened up and the and the justice kept fattened up.

449
00:49:22,450 --> 00:49:25,540
And then the Pentagon has done funny things at the expense of those.

450
00:49:26,290 --> 00:49:29,350
But you can see the relationship to these here.

451
00:49:30,980 --> 00:49:34,280
If there's the the pentacle. There's the pentacle.

452
00:49:34,280 --> 00:49:43,130
There is here and there. So if I add those extra lines, we sort of retrieved a curvilinear version of the one that we had before,

453
00:49:44,600 --> 00:49:52,430
and it would be nice to put those marks on some actual tilings and then tilings easiest ones to make the rhombus ones.

454
00:49:53,480 --> 00:49:58,040
But if you do that, then you find that the rhombus.

455
00:49:58,370 --> 00:50:02,180
Some of them are different. You see that rhombus has a different arrangement in the other.

456
00:50:02,750 --> 00:50:06,710
So we have to add a few more lines. There we go.

457
00:50:07,980 --> 00:50:18,770
And if you do that, then the fat rhombus has everything the same two lines going across each other and the thin rhombus was two lines across.

458
00:50:19,250 --> 00:50:24,140
The pattern itself is now just that.

459
00:50:25,370 --> 00:50:29,510
And what we have first, which is this second.

460
00:50:29,570 --> 00:50:32,900
Let me add a few more lines then we have that arrangement.

461
00:50:33,710 --> 00:50:37,850
And so I was rather hoping that that would be the major feature when you see the

462
00:50:37,850 --> 00:50:43,050
tiles and it was suggested that the tiles should have stainless steel arcs.

463
00:50:43,070 --> 00:50:49,880
I thought that sounded really nice. And I remember coming back, I think I was at a conference in Edinburgh or somewhere up north,

464
00:50:50,060 --> 00:50:54,770
and there were a set of I think six tiles in front of the building.

465
00:50:55,310 --> 00:50:58,790
And I looked at them and there seemed to be two things wrong with them.

466
00:50:59,630 --> 00:51:03,350
One was that the standing still didn't seem to join on.

467
00:51:03,350 --> 00:51:07,070
I mean, they joined, but they but some were bright and some were dark.

468
00:51:07,910 --> 00:51:13,370
And the reason was that they were apparently was called combed and the combing has a different grain to it.

469
00:51:13,880 --> 00:51:17,420
So they said, oh, well, we can fix that by polishing instead of combing it.

470
00:51:18,050 --> 00:51:21,740
But the other problem was that the tiles looked very different.

471
00:51:22,220 --> 00:51:28,610
One was much darker than the other, and that wasn't what I wanted. I want them to look fairly similar to the pattern of the arcs is what you saw.

472
00:51:29,630 --> 00:51:34,370
And then a few days later, I came back and they looked tiles looked the same.

473
00:51:34,520 --> 00:51:38,389
The two different rhombus look the same as each other looks much better.

474
00:51:38,390 --> 00:51:41,570
Maybe they've got some different tiles, but it wasn't that at all.

475
00:51:41,960 --> 00:51:45,710
The reason was the first day I'd seen them it had been raining and they were wet.

476
00:51:46,340 --> 00:51:52,730
And the second day it hadn't. It was nice and dry. And so we thought, well, that's nice because then you get two different kinds of pattern.

477
00:51:52,970 --> 00:51:58,730
When it's dry, you see the mainly the arcs and when it's wet you can bring out the other pattern.

478
00:51:58,740 --> 00:52:04,459
So I thought, that's rather nice. Anyway, here we have the building as a whole.

479
00:52:04,460 --> 00:52:12,290
Well, from that angle. And here we have a sort of shot in front of the main entrance at the back there.

480
00:52:12,860 --> 00:52:21,950
And you can see these rings, the that's the ring of Pentagons as it was now changed into a circle.

481
00:52:22,520 --> 00:52:26,090
And you have several circles and you have sort of bigger ones going round.

482
00:52:27,290 --> 00:52:33,070
Not sure whether you can see it very well here. Hmm. Maybe the other picture is better for that.

483
00:52:37,820 --> 00:52:43,600
Yeah. Here we are. Yeah. So you can see. Let's just check.

484
00:52:45,670 --> 00:52:56,680
I think if you take the circle just up there and then a little bit further out, you'll find a sort of floral pattern of a ten sided shape.

485
00:52:56,990 --> 00:53:05,010
So. I thought it might be amusing just to see, you know, how if you follow the lines of, well, there's a I can give you a conjecture.

486
00:53:05,010 --> 00:53:08,220
I don't know if I want to say it to conjecture or not because I haven't really thought about it much.

487
00:53:08,580 --> 00:53:12,690
But you see these patterns always join up in some way.

488
00:53:13,020 --> 00:53:17,600
But do they form closed loops where you can see some of them?

489
00:53:17,610 --> 00:53:22,050
You feel the circle. Okay, that's a closed loop. You sometimes see double circles.

490
00:53:22,970 --> 00:53:25,170
I'm not sure if I can find one to like that.

491
00:53:25,890 --> 00:53:33,900
That's another shape where you can see the big curvilinear decoration and you also find the curvilinear Pentagon in the right places,

492
00:53:34,410 --> 00:53:39,330
which is something I haven't quite expected. But you see them, too. And so.

493
00:53:41,800 --> 00:53:45,550
You get close friendships. But if they cross each other.

494
00:53:45,580 --> 00:53:49,420
Do you ever get a close shave? Are those four the only ones you ever get?

495
00:53:50,320 --> 00:53:55,900
If you find anything which doesn't belong to him versus just just sort of by inspection.

496
00:53:56,170 --> 00:53:59,650
The only ones I've found, I've never been able to close and they go right off the picture.

497
00:54:00,400 --> 00:54:05,440
So the conjecture might be that those are the only closed loops that may well be false,

498
00:54:05,440 --> 00:54:13,520
because I haven't any idea really this much, because one of them is the shape of the pilings.

499
00:54:15,400 --> 00:54:24,590
If you watch it, they are okay. I think that's one of the few other things I can say, but perhaps people have questions and then to leave it at that.

500
00:54:24,610 --> 00:54:30,640
Oh, I just. Yeah, I'm sure that there's one more picture. No, that's just the beginning, isn't it?

501
00:54:33,090 --> 00:54:37,460
I thought there was one more picture. That's what lost the moment.

502
00:54:37,630 --> 00:54:42,590
Now, it just showed what they look like when they're wet. But. And this is after this?

503
00:54:43,160 --> 00:54:47,720
Mm hmm. No, that's right. At the beginning, I got lost, number one.

504
00:54:48,620 --> 00:54:49,310
Thank you very much.