Tuesday, November 19, 2019

Your Child Has Been Diagnosed With Cancer. Now What?




Three years ago, at the age of 15, my daughter Kate was diagnosed with a rare cancer of the neuroendocrine system, which had started in her pancreas and spread to her liver. It’s one of the most terrifying things a parent can hear.

I’d like to share some of our experiences with you, and pass on some of the things we’ve learned. If your child was recently diagnosed, it will give you an idea of what to expect, and perhaps save you some much needed energy. If a friend or family member is going through something similar, you may find it useful for ideas on how to help.

There are many kinds of cancer, and many can be very effectively treated.

Cancer is not one disease — it is a category with hundreds of specific illnesses. They are all a bit different, with different treatments and outcomes. A cancer diagnosis is scary as hell, but don’t immediately jump to the worst conclusions. It will probably take a few weeks to figure out exactly what kind your child has — after the initial scans or blood work point at cancer, biopsies will likely be done to determine the the type. Your doctor likely won’t be able to tell you much about the specific treatment options until after that.

Do not assume the worst. Some cancers can be kept at bay for a very, very long time with modern
targeted therapies. Some can even be cured.

Ask about genetic testing of the tumor.

Cancer can be caused by mutations to a large number of different genes. Some of those mutations can be targeted by specific treatments that can be remarkably effective. You won’t know which mutations your child’s tumor has without genetic testing. See if your oncologist thinks it is worthwhile, and if they say no, make sure they have a good reason.

Not all oncologists can effectively treat all kinds of cancer.

If the cancer your child has been diagnosed with is rare, ask the physicians making the diagnosis if a specialist would be appropriate. Ask them politely if they feel they are the best qualified to treat your child. The good ones won’t be offended in the least.

It may take some hunting to find the doctor who has the best combination of attitude, skills, and experience to treat your child. They have varying levels of risk that they are willing to take, and you must find someone who matches your views.

An early specialist we spoke to was harsh, negative, and unwilling to take even the slightest risk to aggressively try to cure my child. She wanted to treat Kate like she’d treat a 68 year old. We fired her and found someone who better fit our expectations.

You may have to travel, especially if the cancer is a rare one. Our oncologist is four hours away, but coordinates with a local hospital for supportive care. He and his team are amazing.

While you take care of your child, you must care for yourself, too.

You’re a parent, and your child is in danger. You’d do anything to protect them, and would gladly lay down your life to save them. You can’t, though. It’s not that kind of fight. This is a grind. If you don’t take care of yourself too, you will eventually reach a breaking point.

Your natural impulse is to work feverishly to do whatever might make them more relaxed and comfortable. You need to do that, of course, but make sure to take at least a little time to do the things that help you relax. Get at least a bit of exercise, read, spend a little time with your favorite hobby. At first, you’ll feel guilty when you do — after all, there’s work to be done! Fight that feeling. You are staying strong — you are doing what you need to do to help them fight for years. If you work to failure, you can’t help them as effectively.

When my kid was diagnosed, I bottled it up. I channeled the all the fear into work, and researching doctors and treatment options. I did a terrible job of caring for myself physically and mentally. The anxiety cropped up in weird ways — I developed unreasonable fears of failure at work, selling our house, and other things I actually had a little control over. Don’t do that. As soon as you are able, start a regular routine to keep yourself sane. Your child and the rest of your family depend on you.
Emotionally, you are in for a wild ride. You are not alone. It doesn’t mean you are weak.

You’re likely going to experience a serious roller coaster of emotions — the darkest fears, extreme highs when you get good news, anxiety, depression. That doesn’t mean you’re weak — it means you’re human.

You might well get frustrated at the limitations your child’s illness has placed on your family, and then immediately feel like a terrible person. You aren’t. You are allowed to be frustrated and sad and angry.

Consider talking to a counselor before you feel like you need it. You may well handle it fine in the short term, but again — this is a long haul. You need to prepare for that. Support groups are also an excellent outlet. If you are a church member, they might be a wonderful source of support.
All the terrible stress may trigger depression or anxiety attacks for the first time in your life. You are NOT weak. Get up, get help, and be there for your kid. You can do it. You don’t need to go it alone, and you shouldn’t try.

Keep notes.

Either in physical or virtual form, keep notes. Each time you meet with a doctor, have a procedure, or treatment starts/stops, enter a short summary with the date. It will be invaluable later when talking to providers — you may be surprised at how little different providers actually communicate between themselves. It may also be useful for the inevitable arguments with insurance.

Because we had a blended family, we used a cloud document provider that allowed us to share and edit the document. It worked great.

Lean on your friends and family.

When your friends and family hear that your child is ill, many of them will offer their help in any way they can. They mean it, and at least at first, they have absolutely no idea what sort of help you might need.

You need to ask them. You may have been fiercely independent before, feeling as if asking for help made you weak. Things are different now. You need help.

Be specific. Need someone to mow your grass? Deliver a meal? Drop your other kid at school? Ask them. You’ll be surprised at how many will be thrilled to help you. They really do want to — they just don’t know how.

Sometimes someone trying to comfort you will inadvertently hurt you.

When people become aware that your kid is seriously sick, they feel the need to say something comforting. Society practically demands it. If they don’t have experience with cancer, they will cast about and find something that they hope conveys the right level of concern and hope. Some of them are probably going to make you seriously angry. Expect it, and be ready to roll with it.

My top trigger phrase was “Everything happens for a reason.” That never failed to make my blood boil. You aren’t allowed to hit them. Don’t yell at them. It’s not their fault. They just don’t know what to say. They are trying. Walk away and breathe.

Pack a go-bag and keep it ready.

It is likely that you are going to sometimes go to the hospital on short notice, and stay there for several days. It is extremely helpful to have a bag packed with a change of clothes for you and your child, along with some other essentials. Two cell phone chargers — one for you, one for your child, and an e-book reader or other distraction. Basic toiletries. Caffeine mints can substitute for a cup of coffee in the middle of the night. Any medications you might need, and some common over the counter stuff. You don’t want to be hunting for an ibuprofen or cell phone charger at 3:00 AM.

Consider a private social media group for updates to friends and family.

Very soon, sending texts to update friends and family will get taxing. Be efficient. A Facebook group that interested friends and family can subscribe to allows you to post in one place and broadcast important status information with little effort. Explain to friends that it’s overwhelming to do individual messages. They will immediately get it.

Finances

It will hit you a few days to a week after you get the news, when the first shock wears off. How the hell are you going to pay for this?

You are not alone in this fear — everyone who deals with cancer finds it terrifying, since the numbers are staggering. It will be ok. You will find a way.

Go find the insurance policy that your child is on. Find the section about out-of-pocket maximums. That’s the amount you are going to need to pay per year. It will likely be a pretty large amount of money, but not catastrophic. Once you pay that in co-pays and procedure percentages, it’s likely that you won’t have more medical bills for that year.

What you do need to brace for is the long haul. You are likely going to hit that out of pocket maximum multiple years in a row. You may also have travel, with hotel stays.

Once you know what kind of cancer your child has, look for organizations that support people with that kind of cancer. They may have grants that can help if you need them. Some children’s hospitals reduce fees for those in need. Talk to the social workers that the hospital will introduce you to. If you need to travel, ask if the hospital has a service to help you find hotels at a reduced rate. Many do.

Friends may offer to raise money to help on the internet. Ask them to first look to see if there are any established non-profits that can help with fundraising, since the donations might then be tax deductible for the giver. For-profit fundraising internet sites take a rather substantial portion of the gifts, too. Your hospital social workers may be able to point you at a good option.

Consider a medical ID bracelet.

Cancer patients typically have complicated medical records. Their medical history and medication list can be quite long. Medical ID bracelets are an excellent way to convey critical information to first responders in the event that your child is in an accident, or otherwise unable to communicate. A barcode on the bracelet points at detailed medical records and physician contact information. They are quite inexpensive.

Research treatment options, but don’t dig too deep on outcomes.

Be aware that most of what you read about 5 year survival rates are based on relatively old studies. New treatments are being developed with increasing speed, and that information may well not be current. They are also averages over fairly large groups, so they don’t say a whole lot about the odds of an individual. Many of those studies did not involve children. Every cancer patient is different — stay positive, and prepare for the long fight.

Beware the scammers.

There are people who will try to profit off of your desperate desire to help by selling you “cures” that are not proven to have any benefit whatsoever. They will offer a magic bullet, with anecdotes about how effective they are. These people, in my opinion, are predators who seek you out in your time of gravest need and try to profit from you. They may tell you that the medical and scientific establishment are hiding the cure. That isn’t true — thousands of people working in those industries have family with cancer too, and they would never allow such a secret to be kept. The truth is that curing cancer is really, really hard simply because there are so many kinds.

Keep living. Make memories.

It is very easy to get swept up in the tidal wave of events surrounding an illness as serious as cancer. It takes planning and effort, but you and your family can probably still travel and do other activities you enjoy. There is much more uncertainty involved, and the odds that you’ll need to cancel are much higher than normal, but it’s usually possible. Your doctors will work with you. They understand the importance of staying mentally healthy, and time on vacation or spent with family can help a great deal.

Closing thoughts

This is hard, and it’s completely unfair. Take care of yourself, and ask for whatever help you need. Those around you very much want to help, and will jump at the chance if you let them. Don’t jump to conclusions about outcome, and make sure you get the best medical advice you can. Keep fighting — you don’t know what research is in the works that can help your child. The science is advancing every day, and there are tens of thousands of people fighting beside you. Lean on them, and stay strong.

Please note that these observations are not medical advice. I’m not medically trained in any way — they are based on my own experiences over the last several years. I hope that you find them useful.

Friday, November 15, 2019

The 8 Minute Guide To How Your Business Can Solve Problems with AI and Machine Learning


Ads for machine learning and AI enhanced products are everywhere — if you’ve been to a business conference or talked to a tech vendor in the last two years, you’ve almost certainly heard about it. Promises are everywhere that machine learning and AI will streamline your business, provide valuable insights into your company’s existing data, and improve operations. The promises tend to be a little thin on detail — what, exactly, can it do?

I asked the same question. I present here a concise introduction to what machine learning and AI can do, right now. We’re not going to get bogged down in the math or programming techniques — the goal is to help you see what can be done so you can figure out how it might be applied to your business. 
Let’s get started.

First, a little terminology. Machine learning and deep learning are different methods of problem solving that fall under the umbrella of “artificial intelligence (AI)”. The way these are used are not particularly consistent, especially in marketing literature. For the purposes of this article, the key takeaway for the business person is that deep learning allows solving different kinds of problems than traditional machine learning algorithms, but is generally more expensive and time consuming to implement. It’s also under more active development, and is on the leading edge of what is possible, but you can do a great deal with traditional approaches at less cost.

1) If you have known categories of something, machine learning can sort things into those categories.

This is called classification. You can categorize visitors to your website into buckets based on their click patterns, gaining insights to what they are likely to buy or be interested in. Given some data about a person, a machine learning algorithm can categorize them as likely having an illness or not. Network traffic can be classified as malicious or not. Email can be classified as spam, or not spam. Machines which are likely to fail soon can be identified and maintenance performed prior to failure, based on vibration, current consumption, or other measurements.

Doing this requires a pile of examples that have been sorted into categories already, likely by a human expert. This is the hard and potentially expensive part, and it means that should consider collecting data for future use on a problem like this NOW, even if it’s not immediately clear how you might use it later. You can’t train an algorithm to recognize a failing machine if you don’t have examples of both healthy and not healthy machines, and you may not know that until after a machine fails.

Algorithms that learn from a set of carefully tagged examples are broadly categorized as supervised learning algorithms.

2) If you have a bunch of examples of values, such as what something costs, a machine learning algorithm can learn to predict the value of something that is similar.

An excellent example of this is Zillow’s estimate of what your house is worth if you were to sell it today. Given the number of rooms, square footage, location, number of baths, and other variables, the algorithm makes a prediction based on what similar houses sold for. This is called a regression problem.
There are numerous ways regression problems that might be useful to a business. Given the day of the year and load over the last few days, what load are we likely to see on our server today? How much sunscreen are we likely to sell? What’s a reasonable salary for a person, given experience and a list of skills? Regression can also be used to make financial forecasts.

Like classification, regression is a supervised learning technique and relies on having a significant set of tagged examples that the algorithm can learn from. I think of regression as an “eyeballed” value — a person can take a reasonable stab at it, and get fairly close. The power in machine learning regression is that of scale — it’s easy for a person to make a few dozen such estimates, but a machine can make tens of thousands of such predictions daily. It can be trained on new examples of data in minutes to hours, rather than years.

3) Machine learning can help you find underlying patterns in data that you didn’t know were there.

Supervised learning is powerful, but it has one major limitation — you have to have a good idea what you are looking for, and be able to generate a bunch of carefully tagged examples for the algorithm to learn from.

Algorithms can also be used to look at data and identify clusters of similar entries. This is a kind of unsupervised learning, since it’s not using tagged examples. The classic example is to identify various categories of visitor to your website by demographics and spending patterns.

4) Deep learning can learn to do things by trial and error.

An approach called reinforcement learning can be used to let a computer figure out how to do certain tasks by trying different approaches and seeing what works. A supervisor program assigns a score to each performance based on how well it did, and the learning algorithm makes incremental improvements until goals are met. This approach is useful when even a human expert isn’t sure what the ideal approach is, or where it’s hard to codify the rules. This approach has been used to let multi-legged robots figure out how to walk efficiently and to play video games. For example, I wrote a program that let a computer learn how to land on the moon.

5) Deep learning can find things in pictures and identify them with excellent accuracy. It can also generate text descriptions of a scene.

Deep learning has had a tremendous impact on image recognition. In the last few years it has become fairly easy to train a computer to find and label objects within images. Given a large set of tagged example images, an algorithm is trained, which takes a large amount of computing power. The results are stored as a model, which can be run on much lower power computers. These models are available for download or as software services — you don’t need to train your own unless your images contain something specialized. The models, once trained, can be run on very modest compute devices — they can even be embedded into cameras directly.

Once objects are identified in a video stream, the motion of that object can be tracked. This allows the computer to count people passing through a door, keep track of cars in a parking deck, or other similar tasks. Models are available to isolate and read license plates in an image.

Deep learning models have also been trained to generate descriptions of the overall scene in a picture, with no human input. Models are also being developed to assist radiologists in reading medical imagery, and are getting quite good at doing tasks like finding tumors in CT or MRI scans.

6) Deep learning can generate new images or video.

Deep learning models can be trained to take an image or video and output a modified version of the input. Examples include automatic coloring of black and white images or video frames and generating art. There has also been a great deal of progress on generating fake video streams based on example video. So called “deep fake” videos can generate surprisingly convincing videos of a politician or celebrity giving a speech they never gave. The technology can also generate modified videos where a different person’s face is overlaid on the actor or actress in the original scene in a video.

Deep learning algorithms are also being trained to detect when a video or photograph has been generated in this way, as a countermeasure to potential hostile use of the technology.

7) Machine learning can extract useful information from written text.

Machine learning models performing text analysis (also called natural language processing) can be fed blocks of text and determine if the sentiment is mostly positive or negative. It can be used for monitoring social media or other comments about your brand, product and services, or analyzing other text data for useful business intelligence.

Like other machine learning applications, the important advantages of doing this with a machine are scale, speed, and cost. It becomes possible to examine huge sets of text data and extract useful information far more rapidly than would be possible with a human team.

8) Deep learning can generate fairly good written text.

The OpenAI team recently released GPT2, a deep learning text generation model. Given a few sentences, this model and others like it can generate paragraphs of reasonably coherent text that can appear to have written by a human. GPT2 can be used online to get a feel for what it can do. Pasting in the first few lines of an article from a news site automatically returns a block of text that looks much like a valid news article. The facts are not reliably correct, but it is written in a manner that is fairly believable. The OpenAI team delayed release of GPT2 because of internal concerns that it could be used to generate misleading news articles automatically.

It’s quite possible that models like GPT2 could be used to generate drafts of novels or other books in the near future, perhaps even simulating the styles of established authors. It is unclear what the copyright status of such works would be.

Key points to remember

As you consider the potential ways that machine learning and deep learning can be used in your business, it is useful to keep a few points in mind.

First, these technologies are now quite accessible to people who are not data scientists. You don’t need a large team of specialized software engineers to use them or write them from scratch — you can use cloud software providers with relative ease. If you do want to run it in-house or do custom work, there are excellent libraries that do the heavy lifting for you.

Instead, the effort should be directed at ensuring the quality of your data and determining the best ways to apply the available techniques. Consider where doing tasks that are simple for a human, but in a way that is vastly faster and with increased scope might benefit your business.

Finally, consider that the goal of these technologies need not be to replace humans in the work force. It can be used to can reduce drudgery, improve accuracy and speed, and free them to do more creative tasks. The most successful organizations of the future will likely be those which best adapt to the new capabilities of machines, and blend them with the parts of the enterprise that are uniquely human.


Tuesday, November 5, 2019

The History Of The Telegraph Is Way More Interesting Than You Think


The Cook/Wheatstone Galvanic Telegraph

We've all seen it in books or movies - a guy from the 1870's hunched over a little telegraph key, tapping out messages to someone far off in the distance. It got a sentence or two in my elementary school textbook - Samuel Morse developed the telegraph, and a code to use it. It's a pretty simple device - just an electromagnet pulling down on a spring loaded arm. That's it, right? Simple.

No.

It was much more than that. The development of the practical, commercially viable telegraph spanned more than fifty years and two continents. That elegant little telegraph key and sounder were the result of constant experimentation and refinement. As we came to understand electricity through the work of Volta, Oersted, Henry and others, each advance drove a corresponding improvement in telegraphy. The approaches were ingenious, and looking back, some were pretty amusing.

Morse's electrical telegraph was the eventual winner, but it was far from the first.

The following information and drawn images are sourced from the concisely titled "The telegraph manual : a complete history and description of the semaphoric, electric and magnetic telegraphs of Europe, Asia, Africa, and America, ancient and modern" by Taiaferro Shafner. I ran across it while researching another writing project, and got completely sucked in. This book is more than 800 pages of dense writing and wonderful images - I have selected some of my favorites here. The book is in the public domain, having been published in the US prior to 1924 with no copyright being asserted. The combination of weird technical approaches and the aesthetics of the drawing is just fascinating.

The Early Years: Static Electricity

Though people have likely been shocking themselves after shuffling across carpets since... well... the inventions of carpets, it took the invention of the Leyden jar in 1746 to allow people to collect enough of a charge to be useful. A simple capacitor, the Leyden jar, allowed a charge to be accumulated over time.
Three Leyden jars in series, for increased shock goodness

The charge came from electrostatic generating machines that had been invented several decades before, and the combination allowed experimenters to deliver nasty shocks. One professor Winkler, having shocked the daylights out of himself, proceeded to shock the lady of the house. Twice.


Seriously poor judgement.

The first telegraphs using static electricity transmitted messages by causing pith balls suspended from a fine string to move. This worked, but the machines were fragile, and only demonstrated at close range.

There was, however, one particularly epic design utilizing static electricity. Noting that electricity flowed invisibly within metal foil that was undamaged, but sparked across breaks, Raizen developed what is probably the world's first alphanumeric electronic display in 1794. He took a mesh of foil pathways and carefully cut the outline of letters and numbers into the grid. Each grid, electrically separated from the others, was connected to the sending station by a pair of wires. The resulting bundle of over 70 wires carried electricity to the grid element for each character. The sender of the message could thus energize individual letter grids, and doing so caused sparks to jump across the cut gaps forming the outline of the letter.


Raizen's Telegraph. Sparks arc across the fine gaps, displaying the letter

Imagine that - each letter lighting up in sequence, with a crackling sound, a whiff of ozone,  and a ghostly white/blue glow. It must have seemed like magic then. Consider too - it would have been a LOT less work to just make a bunch of small gaps, each identical, and labelled with the corresponding letter. It would have worked fine - probably even more reliably - but man, Raizen's approach had style.

In 1798, a static electricity telegraph was demonstrated by Dr. Salva, over miles of wire.  No electrostatic telegraph was deployed at significant scale - one can imagine that it was not terribly robust, at least compared to the optical semaphore tower then in common use.

It is impossible to finish the static electricity section without including this image, because it's just awesome.


Exaggerated representation of fright. 

The Battery

In the 1780's, Luigi Galvani figured out that he could make a dead frog's legs twitch by touching them with an assembly made of two types of metals. Alessandro Volta built upon these experiments, sandwiching saline soaked rags with alternating stacks of copper and zinc discs in a stack. This assembly, later known as the voltaic pile, was the first battery that could provide a constant source of electricity.  Volta's experiments were published in 1799.  


Voltaic pile at the Smithsonian American History Museum. Photo by the author.

The basic idea of the voltaic pile was extended and improved into more robust designs by Cruikshank, and later Chester. The  Chester battery, shown below, was used in widespread on commercial telegraph lines in the United States.


The trough battery, by Cruikshank


The Chester battery

The invention and refinement of the practical battery was critical to the later designs of the telegraph, as they were the first source of constant power.

The Electrochemical Telegraph

It isn't clear who first figured out that passing electricity through water would cause it to split into hydrogen and oxygen, making bubbles at the immersed electrode. It appears to have happened very quickly after Volta developed the battery.

By 1809 Samuel von Soemmering had built and demonstrated a functional telegraph using bubbles to indicate the sent letter. It required a separate wire for each encoded character, but cleverly allowed the wires to be configured to send two letters at a time. Soemmering even noted that the volume of hydrogen at one electrode was double that of the production of oxygen at the other, and assigned commonly used letters like vowels to the hydrogen electrodes to make them easier to distinguish.


von Soemmering's bubbly electrochemical telegraph

The author can hardly resist the urge to build a von Soemmering display for an Arduino to read tweets, slowly, with lots of bubbles.

An extremely condensed history of the electromagnet

Other advances were being made in the fledgling science of electricity that would soon render little bubbly wires completely obsolete.

Oersted discovered electromagnetism in 1819, noting that a compass needle moved when passed near wires energized by a voltaic pile. Inn 1820 an experimenter noted that a coil of wire increased the  deflection, and  in 1825 Sturgeon found that wrapping a horseshoe shaped iron bar with coils increased the strength of the resulting electromagnet. Working from 1828 to 1830 Joseph Henry perfected the electromagnet as we know it today - it was he that hit upon the idea of using insulated wire rather than insulating the iron bar. This allowed the coils to be layered on top of each other, further increasing strength. He published the results of his work in 1831. These advances triggered a new race to build a practical telegraph.


Sturgeon's electromagnet, 1825

The Electromagnetic Telegraph

Two primary approaches were used to build telegraphs using electromagnetism. They were invented and developed nearly simultaneously, and commercially viable systems were built using both.

Early on, it was observed that if you passed electricity through a coil of wire around a compass, that the needle would swing. This assembly was called a galvanometer, and it presented a pretty simple way to pass information with electricity. This class of machine was called the needle telegraph. The trick was how to encode information in that needle swing. Several varieties of telegraphs using one or more galvanometers were built and demonstrated - the first, by Pavel Schilling, used a single needle. The needle could be swung right or left by reversing the current, and letters were encoded by the sequence of movement. Right-Left was an A, Right-Right-Right was a B, and so forth. This was perhaps the first binary encoding method of text.

These ideas were further developed into a commercially viable system by Cooke and Wheatstone, which used a very clever encoding method. Patented in 1837, their approach was to use multiple needles, each of which could be swung in either direction by reversing the current. The needles were then used to point at a letter on a table. In the figure below, the needles are indicating the letter "V". This could be done with six wires - one for each galvanometer, and one for the electrical return.


The Cooke-Wheatstone 5-needle telegraph

The other broad approach used an electromagnet to move something besides a needle. This was the space that Samuel Morse and Alfred Vail eventually dominated, but it was an iterative process. Their first machine actually drew the dots and dashes out on strips of paper - an electromagnet was used to pull a marking device on the paper strip as it rolled past. Their original machine is on display at the Smithsonian American History Museum. 


The original Morse/Vail telegraph design

It soon became apparent during tests that it wasn't necessary to write out the message by machine - the operator could distinguish the letters by sound alone. At that point, effort switched to making telegraph "sounders" that emitted sharp clicks. A spring loaded arm was pulled down by an electromagnet to strike a metal point. These were, compared to previous efforts, simple, cheap, and robust. The code that bears Morse's name went through several iterations, becoming more efficient. The most commonly used characters were given the shortest encodings, with the very prevalent "e" being represented by a single short click. 

Original versions of the Morse telegraph used two wires - a signal, and a return. In 1837, a German, Carl von Steinheil, demonstrated that the earth itself could be used in place of the return wire. Only a single strand of copper was needed to send signals for miles. Stringing Morse telegraph wires was thus far less expensive than competitors using more strands, and it was this combination of robust, simple, and relatively cheap that eventually won the race for a commercially viable design. The network spread rapidly throughout the world, even crossing under oceans within a few decades. 


Submerged telegraph cable, 1851 

Like so many other world-altering inventions, the telegraph did not spring fully formed into the mind of a single genius. Rather, it was developed iteratively, with both cooperation and competition between participants spread across two continents. In the end, the solution that won was not the first, or even the easiest to use - it was the solution that was reliable, simple, and inexpensive enough to scale commercially. It relied on the work of dozens of individuals, and was the result of numerous incremental advances. 

The following passage from the book puts it best:

“The discoveries of Volta, Oersted, and Steinheil are to be considered as pre-eminent, in the consummation of the electric telegraph. The first discovered the generating power, the second gave life and strength to that power, when it had become so feeble, that it seemed as though it was struggling in the arms of death; and the latter economized the commercial application of those elements for the uses of man. All telegraphs are formed upon those three discoveries. Let, then, the names of Volta, Oersted, and Steinheil be inscribed in golden capitals upon the bright escutcheon of telegraphic achievements….”

The telegraph manual : a complete history and description of the semaphoric, electric and magnetic telegraphs of Europe, Asia, Africa, and America, ancient and modern by Shaffner, Tal. P. (Taliaferro Preston)

Wednesday, October 16, 2019

A Visit To The National Cryptologic Museum



An hour southwest of Baltimore, MD, a collection of historical treasure is on display in a small, unassuming looking building. The National Cryptologic Museum is a major part of the National Security Agency's public outreach, and the museum is located a short distance from NSA headquarters near Fort Meade. Given the NSA's scope of work and influence, you might expect a rather lavish, modern looking facilty, but that isn't the case. It is quite modest and appears to be partially staffed by volunteers. It's obvious that the dedication of the staff, rather than piles of money, is what keeps the place going. If you're an enthusiast of the history of computers, military history, or especially interested in codes and codebreaking, this little building is worth your time to visit.


Closeup of SIGABA/ECM machine

To get there, you take the exit for the NSA, which, to my amusement, is clearly marked. There's not much else at that exit, and the signs pointing the way to the museum look almost home made. I was greeted by an exceptionally friendly and helpful gentleman, who asked what my interests were, where I was from, and what I hoped to see. I asked a question about how I might use photographs I took there, and he immediately walked me back to the small business office, where the public relations staff member works. As we walked through, I was stunned to see another man, perhaps in his sixties, casually tearing down a 3 rotor Enigma machine on a card table, and remarking that he wasn't sure why it suddenly wasn't working.

I was not at that point aware that the museum maintains a pair of working Enigma machines that the public can touch, and operate. I had assumed that objects of such historic importance would be behind glass. Being able to actually press the keys and watch the encrypted letters light up as the rotors advanced, to feel the force required, was a unique experience I count myself lucky to have had. 

The museum begins with a charming HO scale model railroad town, as part of a display about the encoded symbol language employed by "hobos" in the Depression. A reduced scale model of the Rosetta Stone is positioned near the entrance. It appeared to me that one of the museum's rooms, the Magic room, is sometimes used by Army/government personnel for meetings. They also have STEM outreach programs for the surrounding area with some really neat activities.


Another detail shot of the SIGABA/ECM 

A side room is dedicated to older supercomputers that have been retired by the agency and other modern electronic devices. A Cray XMP-24 and the Harvest robotic tape drive system take up much of the room. A replica of the bugged Great Seal of The United States that was presented to the US ambassador by the Soviet Union in 1945 is shown as well. It contained a novel HF radio transmitter that did not contain a power source - it was only activated by a strong source of electromagnetic energy when needed, and was thus very difficult to detect. A KL7 electromechanical cryptography machines from the fifties and sixties, as well as a KG-57 satellite encryptor from the mid-seventies. 



European cipher disk, used by Denmark 1910-1914

There are some very old manuscripts and some older cipher devices, including a Jefferson-style wheel cipher and a beautiful cipher disk used by the government of Denmark from 1910-1914. I was particularly drawn to a striking early rotor machine machined from brass by Edward Hebern prior to 1920. It was the first machine to use electromechanical rotor


Edward Hebern's rotor machine

What I was most excited to see was the World War 2 equipment, since I've been fascinated by it for some time. Several models of Engima are on display, including an early commercial model, a 3 rotor machine out of it's case, a 4 rotor machine, and a special example of a type believed to be used by Hitler.


Enigma keyboard, plugboard, and rotors


Enigma plugboard detail


Enigma rotor - exploded view

Additionally, the museum has an impressive US Navy cryptographic bombe machine on display. It has a date of Sept 1943 on one part of the chassis, and used a combination of mechanical rotors and vacuum tubes. The bombes functioned as parallel Enigma analogues that could be run through large numbers of possible combinations looking for possible matches. The original machines were developed by a team led by Alan Turing and Tommy Flowers at Bletchley Park in England, and the design transferred to the US. The US machines built on the work of Turing and Flowers, increasing both speed and parallelism. The Navy bombes were used against the 4 rotor naval Enigma.




The bombe machine and closeups of the rotors and tubes

Also on display are German Tunny (SZ40) and Sturgeon cipher machines, as well as a small assembly from the Colossus machine that Tommy Flowers built to decrypt intercepts from them. A final room displays Japanese cipher Red equipment from WW2, and examples of the analogues that were used to break them. There is also a section on the Japanese Purple code, though no intact machines are in the US. A British Typex cipher machine from WW2 is available to examine in the front lobby.

For researchers, a small library is available, and a collection of publications is available for download.

The visitor has the option to accompany a guided tour, or amble about at their own pace. I chose to amble, but listened in on some of the tour. The guide knew the material very well, and appeared genuinely passionate about it. All of the staff I spoke to were professional and helpful. 

The gift shop was interesting - it has a great deal of NSA branded gear including shirts, water bottles, and the like, but also some food products made by someone with a distinct sense of humor.



Thermographic Detection Enhancement Sauce

If you have an interest in any of this equipment and the epic stories of their development, the museum is a must-see. If you are already in the DC area to see the Smithsonian museums, it's a bit over an hour away, and worth it. If, like me, you've been reading about Enigma and Bletchley Park for a while, the opportunity to actually encipher a short message on a real machine is absolutely not to be missed.

You may also find the nearby National Electronics Museum of interest. It was recommended by a staff member at the cryptology museum, and is about a 15 minute drive from there.


All pictures in the post were taken by me at the NSA Cryptologic Museum. They kindly placed no restrictions on publishing images taken within the museum, but did ask that I credit the museum. I thank them for their time and help. If you go and want to take pictures, I recommend bringing a white balance card along, as the ambient lighting is quite warm. 

Saturday, September 21, 2019

The Aviation Writing of Joseph C. Bergling

A photo of Joe taken on the day of his first solo flight.

As I dug through the papers of my great uncle, Joseph C. Bergling, I ran across some hand typed articles he wrote about his aviation experiences in the earliest years of flying. These papers were passed to me by my father before he died. He had received them prior to Joe's death. They are too interesting to stay in a manila folder in my basement, so I have scanned and OCR'd them so that they can be posted here. I hope that someone might find them useful. Copies of these documents have also been given to the Naval Air Museum in Pensacola. 

Joe flew for Civil Air Patrol and worked as a civil servant at an airport for many years. His student pilot's license is signed by a number of aviation pioneers, including Orville Wright. A scan of that document is presented below. (Link to full res scan)





The first is entitled "Near Mishaps Flying". The second appears to be a speech prepared for the OX5 Aviation Pioneers.  Links are to scans of the original typed documents. Converted text follows, so that they will be indexed and searchable.

Smooth air, Uncle Joe.

NEAR MISHAPS FLYING~ WORLD WAR II

By:
Lt. Col. Joseph C. Bergling Civil Air Patrol (Ret.)

During World War II, of over 400 flights I made while on active
duty, there were a number of incidents that were hazardous.
Flying into MAD (Middletown, near Harri s burg, PA), the Towe r
reported wind gusts up to 35 miles per hour, at 160 degrees which was
a 40 degree cross wind on the runway. I approached the runway, which
was 120 degrees, made about a 45 degree left turn, flew along the
Pennsylvania Railroad about as high as the electric poles. Half way
along the field, turned right, about 90 degrees, into the wind.
Starting the final approach, with power on, I headed for the X where
the two runways meet, touching down with the two wheels. When I
tried to bring the tail down the plane took off again, I gunned the
engine, rose to about 15 feet and traveled further across the
intersection of the runways. I throttled back just enough to
maintain flying speed. With no forward motion, I gradually set the
two wheels down near the edge of the runway. I throttled all the way
back, but the tail of the plane was still in the air. This time I
brought the tail down very slowly. As I turned to go toward the
hangar the cross wind banked the plane until the left wing tip nearly
touched the ground. I kicked the right rudder to head into the wind
and stayed there until the tower sent a jeep and two airmen to help
me taxi to the hangar. I think the wind was more than 35 miles per
hour because the stalling speed of the plane was 40 and I was still
flying when the forward motion of the plane had stopped when I was
about 15 feet in the air.

On two occasions at Bolling, I was number one for take-off.
While taxiing to get into the main runway the wind was so strong that
the plane took off before I got on the runway. As the plane left the
ground due to the strong wind, I gave it full throttle, and then
turned to line up with the runway.

One day at Bolling a Consolidated 8-24 "Liberator" had run off
the edge of the runway and damaged its landing gear. I was standing
near the plane when some dark clouds approached the field. It was a
hot day and I noticed a breeze of very cool air. I didn't like what
I felt, very hot and then very cold and the black clouds. I told
some of the men I thought that a strong wind was going to hit us.
With the help of others we untied the Taylorcraft that was on the
ramp and pushed it and a couple of other planes into the hangar.
Just as we closed the door the wind hit. Bolling recorded the wind
above 100 miles per hour. With the rain and wind beating against the
hangar one wondered if it would hold together. Planes bumping into
each other, into the hangars or turning over damaged twenty-seven of
them. A few civilian maintenance men jumped into planes that were
running loose and were able to save them by holding them into the
wind until the storm was over.

A Civil Air Patrol plane that was stationed on the field for
''Tracking Mission" was one of the unlucky ones. Two of the C.A.P.
pilots were untying their plane when the wind struck. When they saw
it was too late they jumped into the plane, the wind blew the plane
over on its back with the two men in it. They were not seriously
hurt.

I had a great respect for thunder storms, knowing what they
could do to light airplanes. I never flew through one, if I couldn't
get around one I would turn back. During the thunderstorm season of
July and August 1943, I flew every day and only had to turn back two
or three times.

Taxiing behind large planes when their engines were revving-up
was always a danger with our small planes. Once when I was taxiing
on the ramp at Bolling a 8-17 ''Flying Fortress" was idling its
engines. Thinking it was safe to cross behind the big plane I
started to taxi. Just as I got behind the big plane the engines
began to rev-up, blowing my plane over until the right wing nearly
touched the ground. I kicked left rudder and headed into the wind
until the four engines were idled. One other time I got behind the
prop wash of a large plane; this time at Martin's Field in Baltimore.
Landing east and over the long runway, flying lower than the top of
the hangars, the plane banked into an almost vertical position. I
gunned the throttle and landed further down the runway ok. Checking
back at operations I found that a Martin B-26 "Marauder" was running
its engine full-throttle with the prop wash crossing the runway.
All flights weren't for parts alone, some times it was to carry
someone someplace or a combination of the two. The shortest trip was
from Bolling to Washington National with Colonel Walter E. Nicol our
Commanding Officer of the Maintenance Section at Bolling. We took
off at Bolling heading toward National, crossed the Potomac River and
landed straight ahead without making a turn. That flight must have
been about two or three minutes at the most. The longest trip was to
Elkins, West Virginia - 2 hours and 27 minutes.

Sometimes the flights would be to fly Captains, later Majors
Springer or W.E. Dinsmore, test pilots at Maintenance to pick up
planes at different fields to deliver to Bolling. Sometimes I would
fly a mechanic to a field to work on a plane, or fly him back after
he made repairs. Some flights were heads of departments to
Middletown; while I checked on parts he would tend to his business.
On one trip flying back to our home base from Middletown with
the Supervisor of the Instrument Shop, the clouds were very low, so
low that they were covering the top of the hills. Visibility was
unlimited. Approaching Bolling Field I couldn't get to the field
because of the low clouds. Zig-zagging back and forth trying to find
a spot through I decided it wasn't possible, then I spotted the
Pennsylvania Railroad. I followed the railroad that was on low
ground to the Anacostia River, down the river to the Potomac. When
we were over the Potomac between Bolling and National Airport,
Bolling Tower called me by radio and asked if I wanted to land,
"wiggle your wings". It reminded me of the old joke, "Tower do you
hear me wiggle the tower''. I made a 180 degree turn and landed.
During the time that I was zig-zagging trying to find a way to the
field the compass was spinning like mad. My passenger, who was an
expert on instruments but not a pilot, could not understand how I
knew where I was. I wasn't even looking at the compass but living
here for 35 years I was familiar with the area. I was looking at the
ground and low clouds trying to find a way into Bolling. The next
morning I was surprised, to say the least, when I looked in the plane
and saw a complete set of blind flying instruments installed in the
Taylorcraft.

The instruments help a lot in flying the plane, especially in
hazy weather or night flying. There was one time, that I have never
forgotten, stranded at ~iddletown for two days because of bad
weather. The next morning I went to operation, as a matter of fact I
was already there, because I had spent my last dollar for food the
night before and didn't have enough money for a bed at the Visitor
Officer Quarters. That morning the sky was clear and there was a
calm. But there was a heavy fog over the river. The field next to
the river was covered with fog and you could only see about
two-thirds of the way down the field. The people at Operations were
upset because a brand new 2nd Lieutenant and new Weather Officer was
calling the weather "contact". I hurriedly made out a clearance,
Operations cleared it and I got out as quickly as I could before they
closed the field. Having picked up a sailor who was hitch-hiking to
Washington, we took off in the direction that would take us down the
River. When we were about two-thirds down the field the tower asked
to give them my position, I reported "just crossing the end of the
runway". With about 30 hours of blind flying practice in a Link
Trainer and with the Taylorcraft now equipped t~ith blind flying
instruments, I was convinced that it would be no problem, with no
wind, to keep on a straight course that would make it possible to
stay away from the high church steeple near the field since flying
down the river there would be no mountains. Flying a straight course
for about 5 minutes we were still in the fog over the river. I
decided then to make a 90 degree turn to get away from the river. A
couple of minutes later we broke out into a perfect clear day. The
passenger had his eyes "glued to the windshield" trying to see
something through the fog. A few minutes after we were in the clear
sky he fell asleep and didn't wake up until we were a few miles from
Bolling. He said in all seriousness, "that was a long ride". By
that remark I could only guess that the five minutes in the fog must
have been misery for him.

In February 1943 the Army wouldn't supply the C.A.P. planes with
fuel. On one trip I took off about 10 miles from Middletown and
landed at Wilson Field to refuel. I decided to visit my
brother-in-law, Paul Long, and his family who lived in Mechanicsburg.
He later became a Lieutenant in the Navy and served in the Pacific.
After coming out of the Navy he became an architect and built over
100 schools, numerous churches, convents and other buildings around
Harrisburg including the City Hall. In semi-retirement he did a lot
of painting and gave classes in art. After spending the night with
his family he drove me to the airport. Checking the weather the
reports were clear at Baltimore and Washington. It had snowed during
the night at Wilson but the weather was good when I took off. Flying
toward York, Pennsylvania I encountered lowering clouds. Following a
road that lead between two high hills I had to stay below the tops of
the hills to keep out of the clouds. Near York I remember looking
down and could see the face of a boy looking up at the plane. Flying
south over the York Road towards Baltimore the ground was getting
higher and the clouds lower and it was snowing. Snow began ta pile
up on the leading edge of the wing. I knew then I was getting into
trouble. At Shrewsburg, Pennsylvania I spotted a field that looked
like a good place to land. Making a 360 degree turn I headed for the
field. Knowing that ice on the wing could effect the stalling speed,
I had to decide at what speed to land. If I came in too slow I would
drop in, if too fast I would run off the edge of the field. Gliding
in about 60 miles per hour, and only a few feet high as I crossed the
beginning of the field at about 55 MPH. At 50 MPH the plane stalled
making a good landing. I had guessed right. The snow on the wings
did make a difference in the stalling speed. The normal stalling
speed for the plane is 40 MPH. A young airman home on leave helped
with the plane. I phoned Lt. Dinsmore and told him where I was. He
asked if I could get off all right. I told him I could and would as
soon as the weather cleared. This was about 9:30 in the morning.
About 3:00 in the afternoon I called Bolling, they said the weather
was ok. I took off and headed for Bolling. Before I got to
Baltimore I could see that I would fly into the same condition I had
flown into in the morning. This time I could see the long line of
clouds ahead in a N.E. to S.W. direction. I flew S.W. along the
clouds staying a safe distance from them. Nearing Rockville,
Maryland I could hear, over the radio, Bolling closing the field. I
landed at Congressional Airport, now Congressional Plaza, and phoned
Lt. Dinsmore. Col. Nicols sent his chauffeur and car to pick up the
parts. It was three days before I flew the plane to Bolling because
of the high ,..Jind. ''Flying" magazine has a column "I Learned About
Flying From That''. Well I learned from that cold front, never try to
push through one. Even if the weather reports are good.
Another time I "Learned-From-That" While climbing after taking
off from M.A.D., up the river, to get about 1,100 feet so I could
turn south to go to Bolling I noticed a buzzard flying over the river
near the top of the mountain. Near the top of the climb the big bird
soared right in front of the plane. I made a steep turn and dove the
plane towards the river to miss the bird and mountain. As I dove the
bird did too staying in front of the plane for a short while and then
bared off to one side. When I pulled out of the dive we had lost
over 500 feet. From that day on I kept my distance from buzzards.
My passenger didn't say a word all the way home.

Although the Taylorcraft in over 600 hours flying never had a
hard landing and the Continental engine, 75 horse power, never quit,
it was almost shot down twice. Captain J.B. Jones, C.A.P., my
Commander at Langley Field borrowed the plane when his wouldn't
start. He flew over the area that he was assigned to, but since he
had cancelled the f li ght a nd had fail ed to report that he was going
to keep the assignment he was shot at by a ground crew member. Seven
20mm cannon shells burst over his head so close that the bullets
shook the plane. We had a little joke about it after things cooled
down. We said it was all right because the Captain in charge
"apologized", he stopped the trigger happy gurrner 1;Jhen he saw 1,.,ihat
was going on.

May 28, 1943 I flew a mechanic, Albert Manganello from Bolling
to Quantico, Marine Air Base about 30 miles down the Potomac River to
work on an Army plane. Flying back up the ri ver we crossed over the
Potomac close to the east bank at less than 500 feet altitude. As we
passed the Navy station we looked down and s aw an explosion, in a
second or so an e normous boom hit the plane . It shook the plane
violently. When the plane stopped shaking I tried the controls and
told Al Manganello that everything was working ok. Checking with the
foreman of the Armor Section at Bolling he said the fire from the
explosion was probably because we were looking down the mouth of the
barrel of a large gun as it went off. I wondered if Al would ever
fly with me again. He did several times.

Flying at about 1,000 feet a few miles from Quantico a Marine
plane appeared overhead about twenty feet away going in the same
direction. I thought, "what is he trying to do, scare me", well he
did. He was number one to land and I followed behind. As I was on
final approach I noticed he was still traveling fast on the ground.
He ground looped Just before he got to the edge of the field but not
in time to keep the tail from bouncing on rocks damaging the plane.
Then I thought maybe he wasn't buzzing me, but was a "near miss'' that
scared him enough to cause him to make a bad landing.

Another time flying over Porter· Street into Bolling "I Learned
About Flying From That''. There had been an article about the danger
of slipping plane while landing. I was using that method often to
bring the plane down where I wanted it to land. This day I was high
coming over the fence, I slipped to lose altitude when all at once
the plane stalls and began to drop like a rock. I gunned the
throttle, leveled the plane and got flying speed just as the wheels
touched the ground. I was careful about slipping after that.
While making an approach west into Bolling, over Porter Street,
that led to the main gate, I was first to land when I saw a Navy C-45
on my left approaching the Navy field north and next to Bolling. We
were on a collision course. I asked Bolling about the plane on my
left, the tower remarked that they had no control over Navy planes.
I pushed the throttle full forward gaining altitude and went around
again.

There were some other scary moments. Flying from Bolling to
Washington National with Col. Nickel, a Navy C-45 buzzed us when we
were about 20 feet in the air. Col. Nickel said "make your landing".
I had cut off this Navy plane. The plane had to go around again.
Explaining to the supervisor of the tower I was sure I heard that my
number 090 was first to land, he took a minute to check then told me
to be careful after this. Later on I heard a plane being called by
radio Navy 090.

Checking my log book there were flights I don't remember, and I
remember some that are not recorded. The one year and four months
that I, or should I say "We" were on active duty, there was about 600
hours flying time which included carrying about 110 passengers and
hundreds of parts, over 400 flights, averaging about 1 hour and 10
minutes per trip. The shortest about 2 or 3 minutes, the longest 2
hours and 27 minutes. Capacity of plane 12 gallons, used 4 gallons
to hour. Range 3 hours at 2,100 RPM and 97 miles per hour. Only one
trip that didn't accomplish anything.

With the end of active duty April 1944, I flew to Hyde Field to
report to the Commander of the 22nd Tow Target Unit of which I was
attached. Flying the pattern, the turns, the final glide, the touch
down, the run down the runway, turning at the taxi strip a t the
proper speed and taxi to the ramp was the most perfect I ever did, it
was flawless. I never felt more than ever as if I was a part of the
plane and the plane was part of me, we were one, this our final day
together we had to be perfect, and we were. A Civil Air Patrol pilot
came over and remarked, "That is the most perfect landing I ever
saw". After "checking it out" with the commander, I flew the plane
to Congressional Airport and turned it over to the owner, Major
Arthur Hyde, the Maryland Wing Commander. I cried when I left my
reliable companion, this beautiful red and black Taylorcraft, NC
34090 that had served our country so flawlessly.

There was one mission in April 1944 I will never forget. Not a
danger to me but to a six year old girl who was dying of meningitis,
in Childrens Hospital, Washington, D.C .. With no penicillin she would
not live long.

Colonel Jarman, Medical Officer at Bolling Field, issued orders
to release to me penicillin that saved the little girls life. The
little girl was my daughter, Frances Marie Bergling.

Later in 1944 I returned to Bolling as a Civil Service Employee
and worked in the office of Maintenance Control, staying a year and
eight months. My total time at Bolling was three years. When the
active duty flying was over, I put part time in the Civil Air Patrol.
I organized and commanded the College Park Flight, was promoted to a
Group Commander of Maryland Wing. 10 May 1948 the National Capitol
Wing was created, I was chosen the Wing Commander.

The Taylorcraft 34090 and I were accident free, others were not
so lucky. 64 members of Civil Air Patrol died and about 100
airplanes were badly damaged or destroyed.

Speech prepared for OX5 Aviation Pioneers

MR. TILLERY, MEMBERS OF 0X5 AVIATION PIONEERS,

BOB WALLACE SAID IT WOULD BE INTERESTING TO HAVE MEMBERS OF 0X5 TO TELL OF THEIR EXPERIENCES IN AVIATION,
MR,  TILLERY ASKED ME TO GIVE A TALK AT THIS MEETING, THE MEMBERS OF 0X5 KNOW THE HISTORY OF AVIATION WELL,
WHAT l WOULD LIKE TO DO IS TO TELL WHERE I WAS WHEN THE HISTORY OF AVIATION WAS BEING MADE,
10TH FEBRUARY 1908, THE SIGNAL CORP OF THE ARMY GAVE THE WRIGHT BROTHERS A CONTRACT TO BUILD THE ARMY'S FIRST AIRPLANE,
THAT WAS THE DAY I WAS BORN,

My MOTHER SAID WHEN I WAS 4 YEARS OLD l WOULD CLIMB OUT THE SECOND STORY WINDOW AND STAND ON THE PORCH ROOF OF MY GRANDPARENTS'
(LOVELESS) HOUSE IN BERWYN HEIGHTS, AND WATCH THE PLANES FLYING FROM THE COLLEGE PARK AIRPORT.

I REMEMBER AT THE AGE OF 8 AT THE ALTAR OF OUR CHURCH WERE STATUES OF TWO ANGELS, I THOUGHT HOW WONDERFUL IT WOULD BE TO HAVE WINGS AND BE ABLE TO FLY.

IN THE WINTER OF 1916-1917, I WAS PLAYING ON THE ICE OF THE DUCK POND AT SOLDIERS HOME WHEN AN AIRPLANE FLEW OVER. I RAN
ACROSS THE ICE TO GET A BETTER LOOK AT THE PLANE WHEN I STEPPED ON A BOARD WHERE THERE WASN'T ANY ICE, I REMEMBER LOOKING UP AND SEEING ICE ABOVE ME, A MAN SAW ME AND HELPED ME OUT OF THE WATER (I THINK THE PLANE WAS A CURTISS PUSHER)

1917 I MADE MY FIRST JUMP---- WITH AN UMBRELLA.  I JUMPED OFF AN EIGHT FOOT WALL, LANDING HARD TAUGHT ME NOT TO TRY THAT AGAIN,
WHEN MY FAMILY LIVED AT 414-10TH STREET, S.E., WASHINGTONJ D.C. IN THE 1920's l WOULD GO OVER TO BOLLING FIELD AND WATCH THE PLANES
FLY, AT THAT TIMEJ THE ARMY AND NAVY USED THE SAME FIELD. THE ARMY'S BUILDINGS WERE ON THE EAST SIDE OF THE FIELD, THE NAVY'S BUILDINGS WERE ON THE WEST ALONG THE SIDE OF THE ANACOSTIA RIVER WHERE IT FLOWED INTO THE POTOMAC.

I REMEMBER SEEING Lt. AL WILLIAMS DEMONSTRATING INVERTED FLYING, THE OUTSIDE LOOP, AN INVERTED VERTICAL CIRCLE COMPLETELY AROUND THE
FIELD AT VERY LOW ALTITUDE. HE GLIDED UPSIDE DOWN IN HIS FINAL APPROACH AND ROLLED OVER JUST IN TIME TO MAKE A PERFECT LANDING.
ABOUT 1925,I TOOK MY FIRST RIDE IN A CURTISS JENNY, 0X5, AT HOOVER FIELD, VIRGINIA, RUTH BOWLING, LATER MY WIFE, AND I HAD
A RIDE IN A SEAPLANE WITH A HISPANO ENGINE AT ARLINGTON BEACH SEAPLANE, INC,, SOUTH END OF HIGHWAY BRIDGE, I HAD A RIDE IN A BERLINGER-JOYCE AT HOOVER.

ABOUT 1927, - ASH I NGTON AIRPORT, INC, TOOK OVER A 1/4 MILE DIRT
RACE TRACK ACROSS THE COLUMBIA ROAD FROM HOOVER FIELD,

I BOUGHT A HEATH PARASOL, SENT IT BACK WHEN I FOUND OUT I

COULD NOT PUT IT TOGETHER.

HERBERT FAY, PILOT OF A NEW WACO 10 SAID HE WOULD GIVE ME FLYING LESSONS FOR WORK IN HELPING TO MAKE THE OLD RACE TRACK INTO A FLYING FIELD.

ABOUT TWO WEEKS LATERJONE OF THE NEW PLANES TAKING OFF FROM THE FIELD HAD A FORCED LANDING ON A FARM SOUTH OF THE OLD BOLLING
FIELD) WHERE THE NEW BOLLING FIELD IS TODAY. THE PRESIDENT OF THE COMPANY FIRED ME FOR LETTING A WASHINGTON POST PHOTOGRAPHER TAKE
A PICTURE OF THE PLANE, So MY TWO WEEKS WORK ENDED WITH NO FLYING LESSONS. I LATER HAD SOME RIDES AT WASHINGTON AIRPORT THAT I PAID FOR, STEWART REISS GAVE ME ONE HOUR DUAL IN A WACO 10J 1927; PART IN VIRGINIA) WEST OF ALEXANDRIA) VIRGINIA; PART AT COLLEGE PARK,

MARCH 1928 MY BROTHER AND I JOINED A FLYING CLUBJ THE DC, AIR LEGION. PAUL GARBER GAVE THE CLUB A LECTURE ON THE HISTORY OF FLYING,
1929 I JOINED THE CAPITAL AERO CLUB, TOOK PRIVATE PILOT'S GROUND COURSE AT D,C, AIR LEGION, MECHANIC COURSE AT THE AVIATION SCHOOL OF AMERICA, PARACHUTE COURSE AT NAVAL AIR STATION,

MADE ONE JUMP AT LOGAN FIELD) BALTIMORE., MAY 1929, lsT SOLO., COLLEGE PARK AIRPORT) THE 11TH HOURJ 11TH DAY, 11TH MONTH) 11 YEARS AFTER WORLD WAR I IN A TRAVEL AIR, WITH A WORLD WAR CURTISS 90 HP OX5 ENGINE,

JANUARY 29, 1930; PRIVATE PILOTS LICENSE NUMBER 11902, FLIGHT TEST AT HOOVER FIELD,
1930; GLIDER LICENSE NUMBER 436,

WHEN LINDBERGH ARRIVED IN WASHINGTON AFTER HIS NEW YORK TO PARIS FLIGHT, I SAW HIM RIDING IN A CONVERTIBLE ON 10TH STREET BETWEEN
F & G.

MY WIFE SAW HIM A NUMBER OF TIMES AT GEORGETOWN UNIVERSITY WHEN HE WAS DOING RESEARCH, AIR MAIL PILOT AT BOLLING, MARCH 4, 1929. THE INAUGURATION PARADE OF PRES1DENT HERBERT HOOVER, I WAS MARCHING WITH THE DC NATIONAL GUARD. WE HAD JUST MARCHED PAST THE PEACE CROSS ON PENNSYLVANIA AVENUE, WHEN I HEARD ENGINES IN THE SKY. THE CLOUDS COMPLETELY COVERED THE SKY, NOT MUCH HIGHER THAN THE CAPITOL DOME, VISIBILITY BELOW THE CLOUDS WAS UNLIMITED. 

I LOOKED UP AND SAW COMING OUT OF THE CLOUDS THE NAVY'S ZEPPELIN, THE Los ANGELES, IT CAME OUT OF THE CLOUDS, NW, OF THE CAPITOL BUILDING NOT MORE THAN A HALF BLOCK FROM THE CAPITOL GROUNDS.

I WAS AMAZED, HOW DID THEY KNOW WHEN TO LET DOWN THROUGH THE CLOUDS AND MISS THE CAPITOL BY ONLY ABOUT A BLOCK OR SO?

THE 800 FOOT ZEPPELIN FLEW ABOVE THE PARADE ROUTE ON PENNSYLVANIA AVENUE, JUST A LITTLE HIGHER THAN THE BUILDINGS,
I REMEMBER THE GOODYEAR BLIMP AT HOOVER, I THINK IT HAD A HANGER THERE. WHILE FLYING AT COLLEGE PARK, l SAW A PLANE THAT WAS EXPERIMENTING WITH BLIND FLYING INSTRUMENTS. THE PLANE WOULD FLY INTO THE FIELD JUST A FEW FEET OFF THE GROUND AND THEN CLIMB AGAIN. I WAS TOLD THAT JIMMY DciOLITTLE WAS ONE OF THE PILOTS.

ONE DAY AT BOLLING FIELD, I DON'T REMEMBER WHAT THE OCCASION WAS, BUT 1,100 AIR PLANES FLEW OVER WASHINGTON.

1928 THE INTERNATIONAL CIVIL AERONATUICAL MEET WAS HELD IN WASHINGTON, AVIATION PEOPLE FROM ALL OVER THE WORLD ATTEND. I WAS ABLE TO GET AUTOGRAPHS ON MY STUDENT PILOT'S LICENSE, OF ORVILLE WRIGHT, LT. AL WILLIAMS, LT. LESTER MAITLAND, CLARENCE D. CHAMBERLIN, FOKKER, LADY MARY HEATH, PARKER CRAMER AND A FEW OTHERS.

ALL THESE EVENTS WERE COVERED BY THE NEWSPAPER RADIO AND MOVING PICTURE NEWS REELS WHICH WAS SEEN AND HEARD WITH GREAT INTEREST.
ERNIE PYLE, THE MOST POPULAR REPORTER IN WORLD WAR ll, HAD A AVIATION COLUMN IN THE WASHINGTON DAILY NEWS ABOUT 1928 TO 1933)
MAYBE LONGER. HIS REPORTING WAS OF GREAT INTEREST IN NEWS OF THE AIR MAIL, BOLLING FIELD NAVAL AIR STATION, WASHINGTON-HOOVER FIELD.

THERE WAS NEWS OF THE SMALL AIRPORTS THAT SPRANG UP IN THE WASHINGTON AREA. COLLEGE PARK, OF COURSE, WAS THE FIRST AND OLDEST, THEN CAME CONGRESSIONAL, SCHROM, QUEENS CHAPEL, CAPITAL, HYDE IN MARYLAND. IN VIRGINIA, BEACON, HYBLA VALLEY AND OTHERS. ALSO NATIONAL NEWS. 

ERNIE PYLE WROTE A LOT ABOUT GLIDING. ONE DAY I GAVE HIM SOME INSTRUCTION IN A GLIDER. HE MADE SEVERAL SOLO FLIGHTS.

ONE DAY AT HOOVER FIELD, ERNIE PYLE AND l HAD A FLIGHT IN A PITCAIRN AUTOGRO, PILOT BY JIM RAY. THE DAY BEFORE HE DEMONSTRATED THE AUTOGYRO, IN FRONT OF THE CAPITOL, FOR MEMBERS OF CONGRESS, ALSO LANDED AT THE WHITE HOUSE GROUNDS,


0N OUR FLIGHT WE TOOK OFF AT HOOVER FIELD, CLIMBED OVER THE AIRPORT AND THE COLUMBIA ROAD THAT SEPARATED THE TWO FIELDS. WHEN WE WERE OVER THE CENTER OF WASHINGTON AIRPOR THE PILOT STARTED THE DESCENT, fROM l000 FEET, WE CAME STRAIGHT DOWN AND LANDED IN THE MIDDLE OF THE FIELD.

1930 I WAS THE INSTRUCTOR OF THE EAGLE GLIDER CLUB. THE 25 MEMBERS MADE ABOUT 400 FLIGHTS. I MADE ABOUT 150 FLIGHTS. RUTH BOWLING, MY FIANCE FLEW THE GLIDER.

JANUARY ll, 1931 I FLEW A RUBBER GLIDER FOR MOVING PICTURE NEWS CAMERAS. THEY TOOK PICTURES OF THE SEVENTH AND EIGHTH FLIGHT, THE GLIDER CRASHED ON THE EIGHTH FLIGHT. THE MOVING PICTURES WERE SHOWN IN THE THEATERS ALL OVER THE COUNTRY. THE EIGHTH FLIGHT, THE CRASH WAS SHOWN IN THE MOVING PICTURE, "THOSE MAGNIFICENT MEN IN THEIR FLYING MACHINES".

BEFORE T.V, WHICH WAS BEFORE WW ll, THE THEATERS WOULD SHOW MOVIES OF NEWS EVENTS ON THE SCREEN. THERE WAS ONE THEATER, TRANS LUX, THAT SHOWED ONLY NEWS EVENTS THAT RUN ABOUT ONE HOUR. IN MAY 193 WHEN THE GERMAN ZEPPELIN HINDENBURG CRASHED, THE TRANS LUX SHOWED THE PICTURES FOR WEEKS,

WHEN OUR COUNTRY ENTERED WORLD WAR II, DECEMBER 1941, I JOINED THE CIVIL AIR PATROL AND WENT ON ACTIVE DUTY FROM NOVEMBER 1942 TO APRIL 1944l STATIONED AT LANGLEY AND BOLLING FIELDS. I WAS A CIVIL SERVICE EMPLOYEE AT BOLLING FROM OCTOBER 1944 TO JUNE 1946.
I WAS IN CIVIL AIR PATROL ABOUT 15 YEARS.

I FLEW PLANES WITH 0X5 ENGINES, WACO 9 & 10, AND TRAVEL AIR ALSO OXX6 LINCOLN PAGE KITTY HAWKJ 100 HP FIVE CYLINDER AIR COOL ENGINE.

THE PIPER CUBS 40, 50, 55 AND 65 HP ENGINES. PIPER CRUISER 75 HP, STINSTON VOYAGER, 75 HP.

PT-17 WITH 225 HP.

CESSNAS 150-152 & 170,

1957 AT A CIVIL AIR PATROL CADET ENCAMPMENT, I HAD A FLIGHT IN A JET T-331  A TWO PLACE F-80. THE PILOT LET ME ROLL THE PLANE AND PERFORM SOME OTHER MINOR MANEUVERS AT NEARLY 500 MILES PER HOUR,