You are on page 1of 6

Comments on Some Conjectures By Amarnath Murthy

Charles Ashbacher
Charles Ashbacher Technologies
Box 294
Hiawatha, IA 52233
cashbacher@yahoo.com

Jeffrey Sand
Student
Mount Mercy College
Cedar Rapids, Iowa

In his forthcoming book [1], Amarnath Murthy defines the Smarandache


Forward Reverse Sum Sequence as

Tn+1 = Tn + R(Tn)

where R(Tn) is the number where the digits of Tn are reversed. The initial
elements of the sequence are

1, 2, 4, 8, 16, 77, 154, 605, 1111, 2222, 4444, 8888, 17776, 85547, . . .

He then puts forward the following two conjectures:

1) There are infinitely many palindromes in the sequence.


2) 16 is the only square in the sequence.

Of course, the second conjecture should be qualified with something like


nontrivial, as 1 and 4 are also perfect squares.
A program in the language Java was written to test these conjectures. The
standard math library of Java contains the BigInteger class, which allows for
integers of very large size to be stored. The only limitation on the size of the
number is the amount of available memory.

The first run of the program searched for elements of the sequence that are
squares for the first 10,000 elements of the sequence. No additional perfect
squares were found. In checking the element to determine if it is a perfect
square, the program determines the integer square root. Since this value will
always be less than or equal to the element, the difference between the
element and the square of the integer square root will give us a measure of
how close we are to a solution. If this difference exhibits a trend, that will be
evidence regarding the truth of the conjecture.

The program was then rerun and the difference between the value of the
element and the largest perfect square less than or equal to the element was
computed for the first fifty elements. The data appears in figure 1.

Figure 1

Note the rapid increase in the difference for sequence elements with index
greater than 40.
However, by itself, this is not conclusive. This shows that the difference
between the largest perfect square less than the element and the element is
growing, but indicates nothing about the other side. It is possible that the
difference between the smallest perfect square greater than or equal to the
element and the element is shrinking. To test this, the program was modified
to compute these differences. The results are summarized in figure 2.

Figure 2

Note that these differences also exhibit a rapid growth for elements
numbered over forty. While there is an element with a substantial drop in
the difference in figure 2, the value of the difference is 125329763. It is only
the large values of the surrounding differences that make it look smaller than
it is.

Finally, the program was rerun for the elements numbered through 300, and
the results are summarized in figure 3. This simply reinforces the principle
that the differences continue to grow at a very fast rate. Therefore, all
evidence indicates that the Murthy conjecture that there are no perfect
squares after 16 is a correct one.

A program was then written to search for additional palindromes in the


sequence. It was used to examine the first 10,000 elements of the sequence
and the only palindromes found appear in table 1.
The program was then modified to determine the number of positions in the
elements that differ from a palindrome. For example, in the number 112211,
this value would be zero and in the number 102214, the value would be two.
This will give us some idea as to whether the elements in the sequence are
close to being palindromic. The values of these differences for the first 200
terms in the sequence are given in figure 4.

Figure 3

Table 1

Element Number Value


1 1
2 2
3 4
4 8
6 77
9 1111
10 2222
11 4444
16 661166
18 3654563

Figure 4

While there are some differences, the upward trend is clearly upward. The
program was then rerun for the first 500 elements of the sequence and the
results appear in figure 5.
Figure 5

From these figures, the evidence is extremely strong that there are no
additional palindromes in the sequence. Therefore, the conjecture by Murthy
is most likely false.

Reference:

1. Amarnath Murthy, Ideas On Smarandache Notions, Theorems,


Propositions, Problems, Conjectures on Smarandache Partitions,
Functions, Sequences, in Number Theory, Combinatorics and
Geometry, American Research Press, Reheboth, Box 141, NM 87322, USA,
(in press).