**教學難點 4: 源於教具不足**

**7. The Easter Eggs in the Examinations WONG Hang-chi**

**7. The Easter Eggs in the Examinations **

*(b) Find the least value of n such that the sum of the first n terms *
of the sequence is greater than 8 × 10^{18} .

The solution of (a) involves some standard techniques of
*geometric sequences. Let a and r be the 1st term and the *
common ratio respectively. According to the question, we have
*two equations ar*^{2}* = 144 and ar*^{5} = 486 . After solving, we have

*r = 1.5 and a = 64 , provided that the terms in the given sequence *

are all real. Since the calculations are quite simple, we omit the
detailed steps here.
Part (b) of the question deserves our attention. The question
*called for “the sum of the first n terms”, which gave us the clue *
to apply the summation formula of a geometric sequence. We
then set up an inequality

64(1.5* ^{n}* – 1)

1.5 – 1 > 8 × 10^{18}
Therefore, we have

1.5* ^{n}* > 6.25 × 10

^{16}+ 1

Since the common logarithmic function is increasing, we have
log 1.5* ^{n}* > log(6.25 × 10

^{16}+ 1)

which implies

*n log 1.5 > log(6.25 × 10*

^{16}+ 1)

*Using a calculator, we obtain n > 95.38167941, and so the *

Everything seems to be fine, except for the step that we apply the logarithm. According to the examination regulations, only calculators that are “HKEAA Approved” may be used in the HKDSE. However, these calculators, such as Casio fx-50FH II, usually do not support high-precision arithmetic (HPA). In fact, Casio fx-50FH II can handle at most 15 significant figures in its memory. In other words, the expression

6.25 × 10^{16} + 1 = 62 500 000 000 000 001

will be rounded off to 62 500 000 000 000 000, correct to 15 significant figures in the calculator memory. For this question, the round off error is tolerable. It is because we can later verify that

64(1.5^{95} – 1)

1.5 – 1 ≈ 6.852981824 × 10^{18} < 8 × 10^{18}
while

64(1.5^{96} – 1)

1.5 – 1 ≈ 1.027947274 × 10^{19} > 8 × 10^{18}

Unfortunately, this method does not always work. It might fail in some situations where the round off error cannot be ignored.

Let us approach this problem in the general situation. Suppose
*that we have a geometric sequence with the 1st term a > 0 and *
*common ratio r > 1 . We aim to find the least value of n such *
*that the sum of the first n terms of the sequence is greater than *

*some constant b > 0 . Using the method suggested above, we *
have

*a(r*

*– 1)*

^{n}*r – 1 > b *

*and so r*

*>*

^{n}*b(r – 1)*

*a*

+ 1
*Using logarithm, we have n log r > log*

*b(r – 1)*

*a*

+ 1
*Thus, the least value of n is *log

*b(r – 1)*

*a*

+ 1
*log r*.

We observe that if

*b(r – 1)*

*a*

is very large, say, over the order of
10^{15}, then adding 1 to it becomes virtually negligible in the calculator. We may get into trouble if it happens that

*b(r – 1)*

*a*

*< r*

*≤*

^{n}*b(r – 1)* *a*

+ 1
It is because when we compute the value of log

*b(r – 1)*

*a*

+ 1
*log r*

using the calculator, the “approximate” value of log

*b(r – 1)*

*a*

*log r*might be returned instead.

To demonstrate this situation of the round off error, let us consider the following example.

In the geometric sequence 9 , 90 , 900 , ... , find the least value
*of n such that the sum of the first n terms of the sequence is at *
least 10^{16} .

The question is equivalent to finding the smallest positive
*integer n such that *

9 + 90 + 900 + ... + 9(10)* ^{n – 1}* ≥ 10

^{16}

It is obvious that the common ratio of the geometric sequence is 10, we can then establish the inequality

9(10* ^{n}* – 1)

10 – 1 ≥ 10^{16}
Therefore, we have

10* ^{n}* ≥ 10

^{16}+ 1 By trial and error, we can easily see that

10^{16} < 10^{16} + 1
and 10^{17} > 10^{16} + 1

Since the sequence 10* ^{n}* is monotonic increasing, we conclude

*that the required least value of n is 17.*

We shall see an interesting phenomenon if we try to solve
10* ^{n}* ≥ 10

^{16}+ 1

using a calculator instead. Since

*we get n log 10 ≥ log(10*^{16} + 1)

*Astonishingly, we arrive at the result n ≥ 16, which gives the *
least value of 16. How come we have two different solutions to
the same question? It seems that we have done each step under
logical deduction, but why the calculator returns an obviously
erroneous answer? Recall that the HKEAA approved Casio
fx-50FH II can handle at most 15 significant figures. The problem
here is that 10^{16} is far larger than 10^{15}. To be precise, we know
that

10^{16} + 1 = 10 000 000 000 000 001

To store this exact value into the calculator, at least 16
significant figures are required, which exceed the memory
capacity of Casio fx-50FH II. Therefore, the calculator will
automatically round off the number to 10^{16}, and thus give rise to
an incorrect result. In fact, we can use computer programs that
support HPA to find the approximate value of
log(10^{16} + 1). The following value is found by Wolfram Alpha,
an online computational knowledge engine developed by the
company Wolfram Research.

log(10^{16} + 1)

≈ 16.000000000000000043429448190325180593640482375 401514738622407081

Interested readers might visit the Wolfram Alpha website
(https://www.wolframalpha.com/) to explore the engine. We
thus find that log(10^{16} + 1) is slightly larger than 16, but a
calculator just do not have enough significant figures to display
this value to a higher precision.

The moral of this question is that checking the answer is always important. A calculator is a merely an aid when solving problems. Not only should we make suitable use of the calculator, but also justify our answers by appropriate logical reasoning.

**2020-DSE-MATH-CP 1 No. 19 **

*PQRS is a quadrilateral paper card, where PQ = 60 cm, * *PS = 40 cm, *

*PQR = 30°, PRQ = 55° and*

*QPS = 120°. The paper card is held with QR lying on the *
horizontal ground as shown in Figure 1.

Figure 1

*(a) Find the length of RS. *

(b) Find the area of the paper card.

(c) It is given that the angle between the paper card and the horizontal ground is 32°.

*(i) Find the shortest distance from P to the horizontal *
ground.

*(ii) A student claims that the angle between RS and the *
horizontal ground is at most 20°. Is the claim correct?

Explain your answer.

Parts (a) and (b) merely involved some relatively simple techniques in trigonometry and mensuration, and so we shall omit the detailed calculations here. But note that the result of (a),

*P *

*Q *

*R *

*S *

*RS ≈ 16.90879944 cm ≈ 16.9 cm will be used in the later parts *

of the question.
Concerning (c)(i), it is helpful for us to draw some auxiliary lines
in the figure in order to help us understand the situation. It is
*clear that the point G on the horizontal ground closest to P must *
*be vertically below P. In other words, G is the projection of P *
*onto the horizontal ground. We then join G and P. Now, the *
*required distance is GP. It is given that the angle between the *
paper card and the horizontal ground is 32°. This information
*lead us to think of the angle between the two planes PQRS and *

*GQR, with the line of intersection QR. It is natural to drop the *

*foot of perpendicular H from P to QR, and draw the lines GH*

*and HP, as shown in Figure 2 below.*

*P *

*Q *

*R *

*S *

*H *
*G *

*K *

*T *

Thus, we have GHP = 32°. As we know that HP = 60 sin 30°

*= 30 cm , we have GP = HP sin 32° = 30 sin 32° ≈ 15.9 cm. *

But wait! In Mathematics, it is crucial to justify that our answer
is correct. By definition, the two arms of the angle between the
*planes PQRS and GQR must be both perpendicular to QR. By *
construction, we can make sure that PHQ = 90°. Nevertheless,
how do we know that GHQ also equals 90° ?

Actually there is a little trick here, namely, the theorem of three
perpendiculars, which will be added into the revised curriculum,
*as mentioned in Learning Objective 14.8 in the Curriculum and *

*Assessment Guide (Secondary 4 - 6). It is recommended that the *

schools should implement the revised senior secondary
Mathematics curriculum for the Compulsory Part at Secondary
4 to 6 progressively from Secondary 4 with effect from the
school year 2023/24.
*The theorem of three perpendiculars states that if GP is *
*perpendicular to the plane GHQ and * *PHQ = 90°, then *

*GHQ = 90°. The marking scheme produced by the HKEAA *
took this theorem for granted in this question, and did not
include a proof of this fact. However, we found it beneficial for
teachers and students to understand the theorem more

thoroughly. Two different proofs of this theorem will be presented here, the first using the elementary geometric approach, and the second using vectors.

*Let us investigate the first method. As GP is perpendicular to the *
*plane GHQ , GP must also be perpendicular to any straight line *
*on plane GHQ , in particular GQ and GH. By Pythagoras’ *

Theorem, we have

*GP*

^{2}

*+ GQ*

^{2}

*= PQ*

^{2}... (1) and

*GP*

^{2}

*+ GH*

^{2}

*= HP*

^{2}... (2) Also, since PHQ = 90°, we deduce that

*HP*

^{2}

*+ HQ*

^{2}

*= PQ*

^{2}... (3) Substituting equations (1) and (2) into (3), we get

*(GP*^{2}* + GH*^{2}*) + HQ*^{2}* = GP*^{2}* + GQ*^{2}

* GH*

^{2}

*+ HQ*

^{2}

*= GQ*

^{2}

*The GP*^{2} term amazingly canceled out, and using the converse
of Pythagoras’ Theorem, we conclude that GHQ = 90°.

The second proof is very elegant and it involves the techniques
of vectors in the Extended Part Module 2. In the foregoing
**arguments, the dot product of any two vectors a and b is denoted **
**by a · b. The term “inner product” or “scalar product” are also **
sometimes used interchangeably in some textbooks, with the

**notation ‹a, b› standing for a · b. It is well known that a · b = 0 **
**if and only if a and b are orthogonal, that is, either a and b are **
**perpendicular to each other, or at least one of the vectors a and **

**b equals 0. **

*We know that GP is perpendicular to the plane GHQ, and so *